OreX/forge-builds
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

initial commit

Browse Source
This commit is contained in:
Bacruru Sakaguchi
2025-10-21 13:37:07 +07:00
commit 9cd16e276a
1574 changed files with 2675557 additions and 0 deletions
Download Patch File Download Diff File Expand all files Collapse all files

185
extensions-builtin/sd_forge_controlnet/.gitignore vendored Executable file
View File

@@ -0,0 +1,185 @@
# Byte-compiled / optimized / DLL files
__pycache__/
*.py[cod]
*$py.class
# C extensions
*.so
# Distribution / packaging
.Python
build/
develop-eggs/
dist/
downloads/
eggs/
.eggs/
lib/
lib64/
parts/
sdist/
var/
wheels/
share/python-wheels/
*.egg-info/
.installed.cfg
*.egg
MANIFEST
# PyInstaller
# Usually these files are written by a python script from a template
# before PyInstaller builds the exe, so as to inject date/other infos into it.
*.manifest
*.spec
# Installer logs
pip-log.txt
pip-delete-this-directory.txt
# Unit test / coverage reports
htmlcov/
.tox/
.nox/
.coverage
.coverage.*
.cache
nosetests.xml
coverage.xml
*.cover
*.py,cover
.hypothesis/
.pytest_cache/
cover/
# Translations
*.mo
*.pot
# Django stuff:
*.log
local_settings.py
db.sqlite3
db.sqlite3-journal
# Flask stuff:
instance/
.webassets-cache
# Scrapy stuff:
.scrapy
# Sphinx documentation
docs/_build/
# PyBuilder
.pybuilder/
target/
# Jupyter Notebook
.ipynb_checkpoints
# IPython
profile_default/
ipython_config.py
# pyenv
# For a library or package, you might want to ignore these files since the code is
# intended to run in multiple environments; otherwise, check them in:
# .python-version
# pipenv
# According to pypa/pipenv#598, it is recommended to include Pipfile.lock in version control.
# However, in case of collaboration, if having platform-specific dependencies or dependencies
# having no cross-platform support, pipenv may install dependencies that don't work, or not
# install all needed dependencies.
#Pipfile.lock
# poetry
# Similar to Pipfile.lock, it is generally recommended to include poetry.lock in version control.
# This is especially recommended for binary packages to ensure reproducibility, and is more
# commonly ignored for libraries.
# https://python-poetry.org/docs/basic-usage/#commit-your-poetrylock-file-to-version-control
#poetry.lock
# pdm
# Similar to Pipfile.lock, it is generally recommended to include pdm.lock in version control.
#pdm.lock
# pdm stores project-wide configurations in .pdm.toml, but it is recommended to not include it
# in version control.
# https://pdm.fming.dev/#use-with-ide
.pdm.toml
# PEP 582; used by e.g. github.com/David-OConnor/pyflow and github.com/pdm-project/pdm
__pypackages__/
# Celery stuff
celerybeat-schedule
celerybeat.pid
# SageMath parsed files
*.sage.py
# Environments
.env
.venv
env/
venv/
ENV/
env.bak/
venv.bak/
# Spyder project settings
.spyderproject
.spyproject
# Rope project settings
.ropeproject
# mkdocs documentation
/site
# mypy
.mypy_cache/
.dmypy.json
dmypy.json
# Pyre type checker
.pyre/
# pytype static type analyzer
.pytype/
# Cython debug symbols
cython_debug/
# PyCharm
# JetBrains specific template is maintained in a separate JetBrains.gitignore that can
# be found at https://github.com/github/gitignore/blob/main/Global/JetBrains.gitignore
# and can be added to the global gitignore or merged into this file. For a more nuclear
# option (not recommended) you can uncomment the following to ignore the entire idea folder.
#.idea
*.pt
*.pth
*.ckpt
*.bin
*.safetensors
# Editor setting metadata
.idea/
.vscode/
detected_maps/
annotator/downloads/
# test results and expectations
web_tests/results/
web_tests/expectations/
tests/web_api/full_coverage/results/
tests/web_api/full_coverage/expectations/
*_diff.png
# Presets
presets/
# Ignore existing dir of hand refiner if exists.
annotator/hand_refiner_portable

674
extensions-builtin/sd_forge_controlnet/LICENSE Executable file
View File

@@ -0,0 +1,674 @@
GNU GENERAL PUBLIC LICENSE
Version 3, 29 June 2007
Copyright (C) 2007 Free Software Foundation, Inc. <https://fsf.org/>
Everyone is permitted to copy and distribute verbatim copies
of this license document, but changing it is not allowed.
Preamble
The GNU General Public License is a free, copyleft license for
software and other kinds of works.
The licenses for most software and other practical works are designed
to take away your freedom to share and change the works. By contrast,
the GNU General Public License is intended to guarantee your freedom to
share and change all versions of a program--to make sure it remains free
software for all its users. We, the Free Software Foundation, use the
GNU General Public License for most of our software; it applies also to
any other work released this way by its authors. You can apply it to
your programs, too.
When we speak of free software, we are referring to freedom, not
price. Our General Public Licenses are designed to make sure that you
have the freedom to distribute copies of free software (and charge for
them if you wish), that you receive source code or can get it if you
want it, that you can change the software or use pieces of it in new
free programs, and that you know you can do these things.
To protect your rights, we need to prevent others from denying you
these rights or asking you to surrender the rights. Therefore, you have
certain responsibilities if you distribute copies of the software, or if
you modify it: responsibilities to respect the freedom of others.
For example, if you distribute copies of such a program, whether
gratis or for a fee, you must pass on to the recipients the same
freedoms that you received. You must make sure that they, too, receive
or can get the source code. And you must show them these terms so they
know their rights.
Developers that use the GNU GPL protect your rights with two steps:
(1) assert copyright on the software, and (2) offer you this License
giving you legal permission to copy, distribute and/or modify it.
For the developers' and authors' protection, the GPL clearly explains
that there is no warranty for this free software. For both users' and
authors' sake, the GPL requires that modified versions be marked as
changed, so that their problems will not be attributed erroneously to
authors of previous versions.
Some devices are designed to deny users access to install or run
modified versions of the software inside them, although the manufacturer
can do so. This is fundamentally incompatible with the aim of
protecting users' freedom to change the software. The systematic
pattern of such abuse occurs in the area of products for individuals to
use, which is precisely where it is most unacceptable. Therefore, we
have designed this version of the GPL to prohibit the practice for those
products. If such problems arise substantially in other domains, we
stand ready to extend this provision to those domains in future versions
of the GPL, as needed to protect the freedom of users.
Finally, every program is threatened constantly by software patents.
States should not allow patents to restrict development and use of
software on general-purpose computers, but in those that do, we wish to
avoid the special danger that patents applied to a free program could
make it effectively proprietary. To prevent this, the GPL assures that
patents cannot be used to render the program non-free.
The precise terms and conditions for copying, distribution and
modification follow.
TERMS AND CONDITIONS
0. Definitions.
"This License" refers to version 3 of the GNU General Public License.
"Copyright" also means copyright-like laws that apply to other kinds of
works, such as semiconductor masks.
"The Program" refers to any copyrightable work licensed under this
License. Each licensee is addressed as "you". "Licensees" and
"recipients" may be individuals or organizations.
To "modify" a work means to copy from or adapt all or part of the work
in a fashion requiring copyright permission, other than the making of an
exact copy. The resulting work is called a "modified version" of the
earlier work or a work "based on" the earlier work.
A "covered work" means either the unmodified Program or a work based
on the Program.
To "propagate" a work means to do anything with it that, without
permission, would make you directly or secondarily liable for
infringement under applicable copyright law, except executing it on a
computer or modifying a private copy. Propagation includes copying,
distribution (with or without modification), making available to the
public, and in some countries other activities as well.
To "convey" a work means any kind of propagation that enables other
parties to make or receive copies. Mere interaction with a user through
a computer network, with no transfer of a copy, is not conveying.
An interactive user interface displays "Appropriate Legal Notices"
to the extent that it includes a convenient and prominently visible
feature that (1) displays an appropriate copyright notice, and (2)
tells the user that there is no warranty for the work (except to the
extent that warranties are provided), that licensees may convey the
work under this License, and how to view a copy of this License. If
the interface presents a list of user commands or options, such as a
menu, a prominent item in the list meets this criterion.
1. Source Code.
The "source code" for a work means the preferred form of the work
for making modifications to it. "Object code" means any non-source
form of a work.
A "Standard Interface" means an interface that either is an official
standard defined by a recognized standards body, or, in the case of
interfaces specified for a particular programming language, one that
is widely used among developers working in that language.
The "System Libraries" of an executable work include anything, other
than the work as a whole, that (a) is included in the normal form of
packaging a Major Component, but which is not part of that Major
Component, and (b) serves only to enable use of the work with that
Major Component, or to implement a Standard Interface for which an
implementation is available to the public in source code form. A
"Major Component", in this context, means a major essential component
(kernel, window system, and so on) of the specific operating system
(if any) on which the executable work runs, or a compiler used to
produce the work, or an object code interpreter used to run it.
The "Corresponding Source" for a work in object code form means all
the source code needed to generate, install, and (for an executable
work) run the object code and to modify the work, including scripts to
control those activities. However, it does not include the work's
System Libraries, or general-purpose tools or generally available free
programs which are used unmodified in performing those activities but
which are not part of the work. For example, Corresponding Source
includes interface definition files associated with source files for
the work, and the source code for shared libraries and dynamically
linked subprograms that the work is specifically designed to require,
such as by intimate data communication or control flow between those
subprograms and other parts of the work.
The Corresponding Source need not include anything that users
can regenerate automatically from other parts of the Corresponding
Source.
The Corresponding Source for a work in source code form is that
same work.
2. Basic Permissions.
All rights granted under this License are granted for the term of
copyright on the Program, and are irrevocable provided the stated
conditions are met. This License explicitly affirms your unlimited
permission to run the unmodified Program. The output from running a
covered work is covered by this License only if the output, given its
content, constitutes a covered work. This License acknowledges your
rights of fair use or other equivalent, as provided by copyright law.
You may make, run and propagate covered works that you do not
convey, without conditions so long as your license otherwise remains
in force. You may convey covered works to others for the sole purpose
of having them make modifications exclusively for you, or provide you
with facilities for running those works, provided that you comply with
the terms of this License in conveying all material for which you do
not control copyright. Those thus making or running the covered works
for you must do so exclusively on your behalf, under your direction
and control, on terms that prohibit them from making any copies of
your copyrighted material outside their relationship with you.
Conveying under any other circumstances is permitted solely under
the conditions stated below. Sublicensing is not allowed; section 10
makes it unnecessary.
3. Protecting Users' Legal Rights From Anti-Circumvention Law.
No covered work shall be deemed part of an effective technological
measure under any applicable law fulfilling obligations under article
11 of the WIPO copyright treaty adopted on 20 December 1996, or
similar laws prohibiting or restricting circumvention of such
measures.
When you convey a covered work, you waive any legal power to forbid
circumvention of technological measures to the extent such circumvention
is effected by exercising rights under this License with respect to
the covered work, and you disclaim any intention to limit operation or
modification of the work as a means of enforcing, against the work's
users, your or third parties' legal rights to forbid circumvention of
technological measures.
4. Conveying Verbatim Copies.
You may convey verbatim copies of the Program's source code as you
receive it, in any medium, provided that you conspicuously and
appropriately publish on each copy an appropriate copyright notice;
keep intact all notices stating that this License and any
non-permissive terms added in accord with section 7 apply to the code;
keep intact all notices of the absence of any warranty; and give all
recipients a copy of this License along with the Program.
You may charge any price or no price for each copy that you convey,
and you may offer support or warranty protection for a fee.
5. Conveying Modified Source Versions.
You may convey a work based on the Program, or the modifications to
produce it from the Program, in the form of source code under the
terms of section 4, provided that you also meet all of these conditions:
a) The work must carry prominent notices stating that you modified
it, and giving a relevant date.
b) The work must carry prominent notices stating that it is
released under this License and any conditions added under section
7. This requirement modifies the requirement in section 4 to
"keep intact all notices".
c) You must license the entire work, as a whole, under this
License to anyone who comes into possession of a copy. This
License will therefore apply, along with any applicable section 7
additional terms, to the whole of the work, and all its parts,
regardless of how they are packaged. This License gives no
permission to license the work in any other way, but it does not
invalidate such permission if you have separately received it.
d) If the work has interactive user interfaces, each must display
Appropriate Legal Notices; however, if the Program has interactive
interfaces that do not display Appropriate Legal Notices, your
work need not make them do so.
A compilation of a covered work with other separate and independent
works, which are not by their nature extensions of the covered work,
and which are not combined with it such as to form a larger program,
in or on a volume of a storage or distribution medium, is called an
"aggregate" if the compilation and its resulting copyright are not
used to limit the access or legal rights of the compilation's users
beyond what the individual works permit. Inclusion of a covered work
in an aggregate does not cause this License to apply to the other
parts of the aggregate.
6. Conveying Non-Source Forms.
You may convey a covered work in object code form under the terms
of sections 4 and 5, provided that you also convey the
machine-readable Corresponding Source under the terms of this License,
in one of these ways:
a) Convey the object code in, or embodied in, a physical product
(including a physical distribution medium), accompanied by the
Corresponding Source fixed on a durable physical medium
customarily used for software interchange.
b) Convey the object code in, or embodied in, a physical product
(including a physical distribution medium), accompanied by a
written offer, valid for at least three years and valid for as
long as you offer spare parts or customer support for that product
model, to give anyone who possesses the object code either (1) a
copy of the Corresponding Source for all the software in the
product that is covered by this License, on a durable physical
medium customarily used for software interchange, for a price no
more than your reasonable cost of physically performing this
conveying of source, or (2) access to copy the
Corresponding Source from a network server at no charge.
c) Convey individual copies of the object code with a copy of the
written offer to provide the Corresponding Source. This
alternative is allowed only occasionally and noncommercially, and
only if you received the object code with such an offer, in accord
with subsection 6b.
d) Convey the object code by offering access from a designated
place (gratis or for a charge), and offer equivalent access to the
Corresponding Source in the same way through the same place at no
further charge. You need not require recipients to copy the
Corresponding Source along with the object code. If the place to
copy the object code is a network server, the Corresponding Source
may be on a different server (operated by you or a third party)
that supports equivalent copying facilities, provided you maintain
clear directions next to the object code saying where to find the
Corresponding Source. Regardless of what server hosts the
Corresponding Source, you remain obligated to ensure that it is
available for as long as needed to satisfy these requirements.
e) Convey the object code using peer-to-peer transmission, provided
you inform other peers where the object code and Corresponding
Source of the work are being offered to the general public at no
charge under subsection 6d.
A separable portion of the object code, whose source code is excluded
from the Corresponding Source as a System Library, need not be
included in conveying the object code work.
A "User Product" is either (1) a "consumer product", which means any
tangible personal property which is normally used for personal, family,
or household purposes, or (2) anything designed or sold for incorporation
into a dwelling. In determining whether a product is a consumer product,
doubtful cases shall be resolved in favor of coverage. For a particular
product received by a particular user, "normally used" refers to a
typical or common use of that class of product, regardless of the status
of the particular user or of the way in which the particular user
actually uses, or expects or is expected to use, the product. A product
is a consumer product regardless of whether the product has substantial
commercial, industrial or non-consumer uses, unless such uses represent
the only significant mode of use of the product.
"Installation Information" for a User Product means any methods,
procedures, authorization keys, or other information required to install
and execute modified versions of a covered work in that User Product from
a modified version of its Corresponding Source. The information must
suffice to ensure that the continued functioning of the modified object
code is in no case prevented or interfered with solely because
modification has been made.
If you convey an object code work under this section in, or with, or
specifically for use in, a User Product, and the conveying occurs as
part of a transaction in which the right of possession and use of the
User Product is transferred to the recipient in perpetuity or for a
fixed term (regardless of how the transaction is characterized), the
Corresponding Source conveyed under this section must be accompanied
by the Installation Information. But this requirement does not apply
if neither you nor any third party retains the ability to install
modified object code on the User Product (for example, the work has
been installed in ROM).
The requirement to provide Installation Information does not include a
requirement to continue to provide support service, warranty, or updates
for a work that has been modified or installed by the recipient, or for
the User Product in which it has been modified or installed. Access to a
network may be denied when the modification itself materially and
adversely affects the operation of the network or violates the rules and
protocols for communication across the network.
Corresponding Source conveyed, and Installation Information provided,
in accord with this section must be in a format that is publicly
documented (and with an implementation available to the public in
source code form), and must require no special password or key for
unpacking, reading or copying.
7. Additional Terms.
"Additional permissions" are terms that supplement the terms of this
License by making exceptions from one or more of its conditions.
Additional permissions that are applicable to the entire Program shall
be treated as though they were included in this License, to the extent
that they are valid under applicable law. If additional permissions
apply only to part of the Program, that part may be used separately
under those permissions, but the entire Program remains governed by
this License without regard to the additional permissions.
When you convey a copy of a covered work, you may at your option
remove any additional permissions from that copy, or from any part of
it. (Additional permissions may be written to require their own
removal in certain cases when you modify the work.) You may place
additional permissions on material, added by you to a covered work,
for which you have or can give appropriate copyright permission.
Notwithstanding any other provision of this License, for material you
add to a covered work, you may (if authorized by the copyright holders of
that material) supplement the terms of this License with terms:
a) Disclaiming warranty or limiting liability differently from the
terms of sections 15 and 16 of this License; or
b) Requiring preservation of specified reasonable legal notices or
author attributions in that material or in the Appropriate Legal
Notices displayed by works containing it; or
c) Prohibiting misrepresentation of the origin of that material, or
requiring that modified versions of such material be marked in
reasonable ways as different from the original version; or
d) Limiting the use for publicity purposes of names of licensors or
authors of the material; or
e) Declining to grant rights under trademark law for use of some
trade names, trademarks, or service marks; or
f) Requiring indemnification of licensors and authors of that
material by anyone who conveys the material (or modified versions of
it) with contractual assumptions of liability to the recipient, for
any liability that these contractual assumptions directly impose on
those licensors and authors.
All other non-permissive additional terms are considered "further
restrictions" within the meaning of section 10. If the Program as you
received it, or any part of it, contains a notice stating that it is
governed by this License along with a term that is a further
restriction, you may remove that term. If a license document contains
a further restriction but permits relicensing or conveying under this
License, you may add to a covered work material governed by the terms
of that license document, provided that the further restriction does
not survive such relicensing or conveying.
If you add terms to a covered work in accord with this section, you
must place, in the relevant source files, a statement of the
additional terms that apply to those files, or a notice indicating
where to find the applicable terms.
Additional terms, permissive or non-permissive, may be stated in the
form of a separately written license, or stated as exceptions;
the above requirements apply either way.
8. Termination.
You may not propagate or modify a covered work except as expressly
provided under this License. Any attempt otherwise to propagate or
modify it is void, and will automatically terminate your rights under
this License (including any patent licenses granted under the third
paragraph of section 11).
However, if you cease all violation of this License, then your
license from a particular copyright holder is reinstated (a)
provisionally, unless and until the copyright holder explicitly and
finally terminates your license, and (b) permanently, if the copyright
holder fails to notify you of the violation by some reasonable means
prior to 60 days after the cessation.
Moreover, your license from a particular copyright holder is
reinstated permanently if the copyright holder notifies you of the
violation by some reasonable means, this is the first time you have
received notice of violation of this License (for any work) from that
copyright holder, and you cure the violation prior to 30 days after
your receipt of the notice.
Termination of your rights under this section does not terminate the
licenses of parties who have received copies or rights from you under
this License. If your rights have been terminated and not permanently
reinstated, you do not qualify to receive new licenses for the same
material under section 10.
9. Acceptance Not Required for Having Copies.
You are not required to accept this License in order to receive or
run a copy of the Program. Ancillary propagation of a covered work
occurring solely as a consequence of using peer-to-peer transmission
to receive a copy likewise does not require acceptance. However,
nothing other than this License grants you permission to propagate or
modify any covered work. These actions infringe copyright if you do
not accept this License. Therefore, by modifying or propagating a
covered work, you indicate your acceptance of this License to do so.
10. Automatic Licensing of Downstream Recipients.
Each time you convey a covered work, the recipient automatically
receives a license from the original licensors, to run, modify and
propagate that work, subject to this License. You are not responsible
for enforcing compliance by third parties with this License.
An "entity transaction" is a transaction transferring control of an
organization, or substantially all assets of one, or subdividing an
organization, or merging organizations. If propagation of a covered
work results from an entity transaction, each party to that
transaction who receives a copy of the work also receives whatever
licenses to the work the party's predecessor in interest had or could
give under the previous paragraph, plus a right to possession of the
Corresponding Source of the work from the predecessor in interest, if
the predecessor has it or can get it with reasonable efforts.
You may not impose any further restrictions on the exercise of the
rights granted or affirmed under this License. For example, you may
not impose a license fee, royalty, or other charge for exercise of
rights granted under this License, and you may not initiate litigation
(including a cross-claim or counterclaim in a lawsuit) alleging that
any patent claim is infringed by making, using, selling, offering for
sale, or importing the Program or any portion of it.
11. Patents.
A "contributor" is a copyright holder who authorizes use under this
License of the Program or a work on which the Program is based. The
work thus licensed is called the contributor's "contributor version".
A contributor's "essential patent claims" are all patent claims
owned or controlled by the contributor, whether already acquired or
hereafter acquired, that would be infringed by some manner, permitted
by this License, of making, using, or selling its contributor version,
but do not include claims that would be infringed only as a
consequence of further modification of the contributor version. For
purposes of this definition, "control" includes the right to grant
patent sublicenses in a manner consistent with the requirements of
this License.
Each contributor grants you a non-exclusive, worldwide, royalty-free
patent license under the contributor's essential patent claims, to
make, use, sell, offer for sale, import and otherwise run, modify and
propagate the contents of its contributor version.
In the following three paragraphs, a "patent license" is any express
agreement or commitment, however denominated, not to enforce a patent
(such as an express permission to practice a patent or covenant not to
sue for patent infringement). To "grant" such a patent license to a
party means to make such an agreement or commitment not to enforce a
patent against the party.
If you convey a covered work, knowingly relying on a patent license,
and the Corresponding Source of the work is not available for anyone
to copy, free of charge and under the terms of this License, through a
publicly available network server or other readily accessible means,
then you must either (1) cause the Corresponding Source to be so
available, or (2) arrange to deprive yourself of the benefit of the
patent license for this particular work, or (3) arrange, in a manner
consistent with the requirements of this License, to extend the patent
license to downstream recipients. "Knowingly relying" means you have
actual knowledge that, but for the patent license, your conveying the
covered work in a country, or your recipient's use of the covered work
in a country, would infringe one or more identifiable patents in that
country that you have reason to believe are valid.
If, pursuant to or in connection with a single transaction or
arrangement, you convey, or propagate by procuring conveyance of, a
covered work, and grant a patent license to some of the parties
receiving the covered work authorizing them to use, propagate, modify
or convey a specific copy of the covered work, then the patent license
you grant is automatically extended to all recipients of the covered
work and works based on it.
A patent license is "discriminatory" if it does not include within
the scope of its coverage, prohibits the exercise of, or is
conditioned on the non-exercise of one or more of the rights that are
specifically granted under this License. You may not convey a covered
work if you are a party to an arrangement with a third party that is
in the business of distributing software, under which you make payment
to the third party based on the extent of your activity of conveying
the work, and under which the third party grants, to any of the
parties who would receive the covered work from you, a discriminatory
patent license (a) in connection with copies of the covered work
conveyed by you (or copies made from those copies), or (b) primarily
for and in connection with specific products or compilations that
contain the covered work, unless you entered into that arrangement,
or that patent license was granted, prior to 28 March 2007.
Nothing in this License shall be construed as excluding or limiting
any implied license or other defenses to infringement that may
otherwise be available to you under applicable patent law.
12. No Surrender of Others' Freedom.
If conditions are imposed on you (whether by court order, agreement or
otherwise) that contradict the conditions of this License, they do not
excuse you from the conditions of this License. If you cannot convey a
covered work so as to satisfy simultaneously your obligations under this
License and any other pertinent obligations, then as a consequence you may
not convey it at all. For example, if you agree to terms that obligate you
to collect a royalty for further conveying from those to whom you convey
the Program, the only way you could satisfy both those terms and this
License would be to refrain entirely from conveying the Program.
13. Use with the GNU Affero General Public License.
Notwithstanding any other provision of this License, you have
permission to link or combine any covered work with a work licensed
under version 3 of the GNU Affero General Public License into a single
combined work, and to convey the resulting work. The terms of this
License will continue to apply to the part which is the covered work,
but the special requirements of the GNU Affero General Public License,
section 13, concerning interaction through a network will apply to the
combination as such.
14. Revised Versions of this License.
The Free Software Foundation may publish revised and/or new versions of
the GNU General Public License from time to time. Such new versions will
be similar in spirit to the present version, but may differ in detail to
address new problems or concerns.
Each version is given a distinguishing version number. If the
Program specifies that a certain numbered version of the GNU General
Public License "or any later version" applies to it, you have the
option of following the terms and conditions either of that numbered
version or of any later version published by the Free Software
Foundation. If the Program does not specify a version number of the
GNU General Public License, you may choose any version ever published
by the Free Software Foundation.
If the Program specifies that a proxy can decide which future
versions of the GNU General Public License can be used, that proxy's
public statement of acceptance of a version permanently authorizes you
to choose that version for the Program.
Later license versions may give you additional or different
permissions. However, no additional obligations are imposed on any
author or copyright holder as a result of your choosing to follow a
later version.
15. Disclaimer of Warranty.
THERE IS NO WARRANTY FOR THE PROGRAM, TO THE EXTENT PERMITTED BY
APPLICABLE LAW. EXCEPT WHEN OTHERWISE STATED IN WRITING THE COPYRIGHT
HOLDERS AND/OR OTHER PARTIES PROVIDE THE PROGRAM "AS IS" WITHOUT WARRANTY
OF ANY KIND, EITHER EXPRESSED OR IMPLIED, INCLUDING, BUT NOT LIMITED TO,
THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR
PURPOSE. THE ENTIRE RISK AS TO THE QUALITY AND PERFORMANCE OF THE PROGRAM
IS WITH YOU. SHOULD THE PROGRAM PROVE DEFECTIVE, YOU ASSUME THE COST OF
ALL NECESSARY SERVICING, REPAIR OR CORRECTION.
16. Limitation of Liability.
IN NO EVENT UNLESS REQUIRED BY APPLICABLE LAW OR AGREED TO IN WRITING
WILL ANY COPYRIGHT HOLDER, OR ANY OTHER PARTY WHO MODIFIES AND/OR CONVEYS
THE PROGRAM AS PERMITTED ABOVE, BE LIABLE TO YOU FOR DAMAGES, INCLUDING ANY
GENERAL, SPECIAL, INCIDENTAL OR CONSEQUENTIAL DAMAGES ARISING OUT OF THE
USE OR INABILITY TO USE THE PROGRAM (INCLUDING BUT NOT LIMITED TO LOSS OF
DATA OR DATA BEING RENDERED INACCURATE OR LOSSES SUSTAINED BY YOU OR THIRD
PARTIES OR A FAILURE OF THE PROGRAM TO OPERATE WITH ANY OTHER PROGRAMS),
EVEN IF SUCH HOLDER OR OTHER PARTY HAS BEEN ADVISED OF THE POSSIBILITY OF
SUCH DAMAGES.
17. Interpretation of Sections 15 and 16.
If the disclaimer of warranty and limitation of liability provided
above cannot be given local legal effect according to their terms,
reviewing courts shall apply local law that most closely approximates
an absolute waiver of all civil liability in connection with the
Program, unless a warranty or assumption of liability accompanies a
copy of the Program in return for a fee.
END OF TERMS AND CONDITIONS
How to Apply These Terms to Your New Programs
If you develop a new program, and you want it to be of the greatest
possible use to the public, the best way to achieve this is to make it
free software which everyone can redistribute and change under these terms.
To do so, attach the following notices to the program. It is safest
to attach them to the start of each source file to most effectively
state the exclusion of warranty; and each file should have at least
the "copyright" line and a pointer to where the full notice is found.
<one line to give the program's name and a brief idea of what it does.>
Copyright (C) <year> <name of author>
This program is free software: you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by
the Free Software Foundation, either version 3 of the License, or
(at your option) any later version.
This program is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
GNU General Public License for more details.
You should have received a copy of the GNU General Public License
along with this program. If not, see <https://www.gnu.org/licenses/>.
Also add information on how to contact you by electronic and paper mail.
If the program does terminal interaction, make it output a short
notice like this when it starts in an interactive mode:
<program> Copyright (C) <year> <name of author>
This program comes with ABSOLUTELY NO WARRANTY; for details type `show w'.
This is free software, and you are welcome to redistribute it
under certain conditions; type `show c' for details.
The hypothetical commands `show w' and `show c' should show the appropriate
parts of the General Public License. Of course, your program's commands
might be different; for a GUI interface, you would use an "about box".
You should also get your employer (if you work as a programmer) or school,
if any, to sign a "copyright disclaimer" for the program, if necessary.
For more information on this, and how to apply and follow the GNU GPL, see
<https://www.gnu.org/licenses/>.
The GNU General Public License does not permit incorporating your program
into proprietary programs. If your program is a subroutine library, you
may consider it more useful to permit linking proprietary applications with
the library. If this is what you want to do, use the GNU Lesser General
Public License instead of this License. But first, please read
<https://www.gnu.org/licenses/why-not-lgpl.html>.

66
extensions-builtin/sd_forge_controlnet/install.py Executable file
View File

@@ -0,0 +1,66 @@
import launch
import pkg_resources
import sys
import os
import shutil
import platform
from pathlib import Path
from typing import Tuple, Optional
repo_root = Path(__file__).parent
main_req_file = repo_root / "requirements.txt"
def comparable_version(version: str) -> Tuple:
return tuple(map(int, version.split(".")))
def get_installed_version(package: str) -> Optional[str]:
try:
return pkg_resources.get_distribution(package).version
except Exception:
return None
def extract_base_package(package_string: str) -> str:
base_package = package_string.split("@git")[0]
return base_package
def install_requirements(req_file):
with open(req_file) as file:
for package in file:
try:
package = package.strip()
if "==" in package:
package_name, package_version = package.split("==")
installed_version = get_installed_version(package_name)
if installed_version != package_version:
launch.run_pip(
f"install -U {package}",
f"sd-forge-controlnet requirement: changing {package_name} version from {installed_version} to {package_version}",
)
elif ">=" in package:
package_name, package_version = package.split(">=")
installed_version = get_installed_version(package_name)
if not installed_version or comparable_version(
installed_version
) < comparable_version(package_version):
launch.run_pip(
f"install -U {package}",
f"sd-forge-controlnet requirement: changing {package_name} version from {installed_version} to {package_version}",
)
elif not launch.is_installed(extract_base_package(package)):
launch.run_pip(
f"install {package}",
f"sd-forge-controlnet requirement: {package}",
)
except Exception as e:
print(e)
print(
f"Warning: Failed to install {package}, some preprocessors may not work."
)
install_requirements(main_req_file)

403
extensions-builtin/sd_forge_controlnet/javascript/active_units.js Executable file
View File

@@ -0,0 +1,403 @@
/**
* Give a badge on ControlNet Accordion indicating total number of active
* units.
* Make active unit's tab name green.
* Append control type to tab name.
* Disable resize mode selection when A1111 img2img input is used.
*/
(function () {
const cnetAllAccordions = new Set();
onUiUpdate(() => {
const ImgChangeType = {
NO_CHANGE: 0,
REMOVE: 1,
ADD: 2,
SRC_CHANGE: 3,
};
function imgChangeObserved(mutationsList) {
// Iterate over all mutations that just occured
for (let mutation of mutationsList) {
// Check if the mutation is an addition or removal of a node
if (mutation.type === 'childList') {
// Check if nodes were added
if (mutation.addedNodes.length > 0) {
for (const node of mutation.addedNodes) {
if (node.tagName === 'IMG') {
return ImgChangeType.ADD;
}
}
}
// Check if nodes were removed
if (mutation.removedNodes.length > 0) {
for (const node of mutation.removedNodes) {
if (node.tagName === 'IMG') {
return ImgChangeType.REMOVE;
}
}
}
}
// Check if the mutation is a change of an attribute
else if (mutation.type === 'attributes') {
if (mutation.target.tagName === 'IMG' && mutation.attributeName === 'src') {
return ImgChangeType.SRC_CHANGE;
}
}
}
return ImgChangeType.NO_CHANGE;
}
function childIndex(element) {
// Get all child nodes of the parent
let children = Array.from(element.parentNode.childNodes);
// Filter out non-element nodes (like text nodes and comments)
children = children.filter(child => child.nodeType === Node.ELEMENT_NODE);
return children.indexOf(element);
}
function imageInputDisabledAlert() {
alert('Inpaint control type must use a1111 input in img2img mode.');
}
class ControlNetUnitTab {
constructor(tab, accordion) {
this.tab = tab;
this.tabOpen = false; // Whether the tab is open.
this.accordion = accordion;
this.isImg2Img = tab.querySelector('.cnet-mask-upload').id.includes('img2img');
this.enabledAccordionCheckbox = tab.querySelector('.input-accordion-checkbox');
this.enabledCheckbox = tab.querySelector('.cnet-unit-enabled input');
this.inputImage = tab.querySelector('.cnet-input-image-group .cnet-image input[type="file"]');
this.inputImageContainer = tab.querySelector('.cnet-input-image-group .cnet-image');
this.generatedImageGroup = tab.querySelector('.cnet-generated-image-group');
this.maskImageGroup = tab.querySelector('.cnet-mask-image-group');
this.inputImageGroup = tab.querySelector('.cnet-input-image-group');
this.controlTypeRadios = tab.querySelectorAll('.controlnet_control_type_filter_group input[type="radio"]');
this.resizeModeRadios = tab.querySelectorAll('.controlnet_resize_mode_radio input[type="radio"]');
this.runPreprocessorButton = tab.querySelector('.cnet-run-preprocessor');
this.tabs = tab.parentNode;
this.tabIndex = childIndex(tab);
// By default the InputAccordion checkbox is linked with the state
// of accordion's open/close state. To disable this link, we can
// simulate click to check the checkbox and uncheck it.
this.enabledAccordionCheckbox.click();
this.enabledAccordionCheckbox.click();
this.sync_enabled_checkbox();
this.attachEnabledButtonListener();
this.attachControlTypeRadioListener();
this.attachImageUploadListener();
this.attachImageStateChangeObserver();
this.attachA1111SendInfoObserver();
this.attachAccordionStateObserver();
}
/**
* Sync the states of enabledCheckbox and enabledAccordionCheckbox.
*/
sync_enabled_checkbox() {
this.enabledCheckbox.addEventListener("change", () => {
if (this.enabledAccordionCheckbox.checked != this.enabledCheckbox.checked) {
this.enabledAccordionCheckbox.click();
}
});
this.enabledAccordionCheckbox.addEventListener("change", () => {
if (this.enabledCheckbox.checked != this.enabledAccordionCheckbox.checked) {
this.enabledCheckbox.click();
}
});
}
/**
* Get the span that has text "Unit {X}".
*/
getUnitHeaderTextElement() {
return this.tab.querySelector(
`button > span:nth-child(1)`
);
}
getActiveControlType() {
for (let radio of this.controlTypeRadios) {
if (radio.checked) {
return radio.value;
}
}
return undefined;
}
updateActiveState() {
const unitHeader = this.getUnitHeaderTextElement();
if (!unitHeader) return;
if (this.enabledCheckbox.checked) {
unitHeader.classList.add('cnet-unit-active');
} else {
unitHeader.classList.remove('cnet-unit-active');
}
}
updateActiveUnitCount() {
function getActiveUnitCount(checkboxes) {
let activeUnitCount = 0;
for (const checkbox of checkboxes) {
if (checkbox.checked)
activeUnitCount++;
}
return activeUnitCount;
}
const checkboxes = this.accordion.querySelectorAll('.cnet-unit-enabled input');
const span = this.accordion.querySelector('.label-wrap span');
// Remove existing badge.
if (span.childNodes.length !== 1) {
span.removeChild(span.lastChild);
}
// Add new badge if necessary.
const activeUnitCount = getActiveUnitCount(checkboxes);
if (activeUnitCount > 0) {
const div = document.createElement('div');
div.classList.add('cnet-badge');
div.classList.add('primary');
div.innerHTML = `${activeUnitCount} unit${activeUnitCount > 1 ? 's' : ''}`;
span.appendChild(div);
}
}
/**
* Add the active control type to tab displayed text.
*/
updateActiveControlType() {
const unitHeader = this.getUnitHeaderTextElement();
if (!unitHeader) return;
// Remove the control if exists
const controlTypeSuffix = unitHeader.querySelector('.control-type-suffix');
if (controlTypeSuffix) controlTypeSuffix.remove();
// Add new suffix.
const controlType = this.getActiveControlType();
if (controlType === 'All') return;
const span = document.createElement('span');
span.innerHTML = `[${controlType}]`;
span.classList.add('control-type-suffix');
unitHeader.appendChild(span);
}
getInputImageSrc() {
const img = this.inputImageGroup.querySelector('.cnet-image .forge-image');
return (img && img.src.startsWith('data')) ? img.src : null;
}
getPreprocessorPreviewImageSrc() {
const img = this.generatedImageGroup.querySelector('.cnet-image .forge-image');
return (img && img.src.startsWith('data')) ? img.src : null;
}
getMaskImageSrc() {
function isEmptyCanvas(canvas) {
if (!canvas) return true;
if (canvas.width == 0 || canvas.height ==0) return true;
const ctx = canvas.getContext('2d');
// Get the image data
const imageData = ctx.getImageData(0, 0, canvas.width, canvas.height);
const data = imageData.data; // This is a Uint8ClampedArray
// Check each pixel
let isPureBlack = true;
for (let i = 0; i < data.length; i += 4) {
if (data[i] !== 0 || data[i + 1] !== 0 || data[i + 2] !== 0) { // Check RGB values
isPureBlack = false;
break;
}
}
return isPureBlack;
}
const maskImg = this.maskImageGroup.querySelector('.cnet-mask-image .forge-image');
// Hand-drawn mask on mask upload.
const handDrawnMaskCanvas = this.maskImageGroup.querySelector('.cnet-mask-image .forge-drawing-canvas');
// Hand-drawn mask on input image upload.
const inputImageHandDrawnMaskCanvas = this.inputImageGroup.querySelector('.cnet-image .forge-drawing-canvas');
if (!isEmptyCanvas(handDrawnMaskCanvas)) {
return handDrawnMaskCanvas.toDataURL();
} else if (maskImg && maskImg.src.startsWith('data')) {
return maskImg.src;
} else if (!isEmptyCanvas(inputImageHandDrawnMaskCanvas)) {
return inputImageHandDrawnMaskCanvas.toDataURL();
} else {
return null;
}
}
setThumbnail(imgSrc, maskSrc) {
if (!imgSrc) return;
const unitHeader = this.getUnitHeaderTextElement();
if (!unitHeader) return;
const img = document.createElement('img');
img.src = imgSrc;
img.classList.add('cnet-thumbnail');
unitHeader.appendChild(img);
if (maskSrc) {
const mask = document.createElement('img');
mask.src = maskSrc;
mask.classList.add('cnet-thumbnail');
unitHeader.appendChild(mask);
}
}
removeThumbnail() {
const unitHeader = this.getUnitHeaderTextElement();
if (!unitHeader) return;
const imgs = unitHeader.querySelectorAll('.cnet-thumbnail');
for (const img of imgs) {
img.remove();
}
}
/**
* When the accordion is folded, display a thumbnail of input image
* and mask on the accordion header.
*/
updateInputImageThumbnail() {
if (!opts.controlnet_input_thumbnail) return;
if (this.tabOpen) {
this.removeThumbnail();
} else {
this.setThumbnail(this.getInputImageSrc(), this.getMaskImageSrc());
}
}
attachEnabledButtonListener() {
this.enabledCheckbox.addEventListener('change', () => {
this.updateActiveState();
this.updateActiveUnitCount();
});
}
attachControlTypeRadioListener() {
for (const radio of this.controlTypeRadios) {
radio.addEventListener('change', () => {
this.updateActiveControlType();
});
}
}
attachImageUploadListener() {
// Automatically check `enable` checkbox when image is uploaded.
this.inputImage.addEventListener('change', (event) => {
if (!event.target.files) return;
if (!this.enabledCheckbox.checked)
this.enabledCheckbox.click();
});
// Automatically check `enable` checkbox when JSON pose file is uploaded.
this.tab.querySelector('.cnet-upload-pose input').addEventListener('change', (event) => {
if (!event.target.files) return;
if (!this.enabledCheckbox.checked)
this.enabledCheckbox.click();
});
}
attachImageStateChangeObserver() {
new MutationObserver((mutationsList) => {
const changeObserved = imgChangeObserved(mutationsList);
if (changeObserved === ImgChangeType.ADD) {
// enabling the run preprocessor button
this.runPreprocessorButton.removeAttribute("disabled");
this.runPreprocessorButton.title = 'Run preprocessor';
}
if (changeObserved === ImgChangeType.REMOVE) {
// disabling the run preprocessor button
this.runPreprocessorButton.setAttribute("disabled", true);
this.runPreprocessorButton.title = "No ControlNet input image available";
}
}).observe(this.inputImageContainer, {
childList: true,
subtree: true,
});
}
/**
* Observe send PNG info buttons in A1111, as they can also directly
* set states of ControlNetUnit.
*/
attachA1111SendInfoObserver() {
const pasteButtons = gradioApp().querySelectorAll('#paste');
const pngButtons = gradioApp().querySelectorAll(
this.isImg2Img ?
'#img2img_tab, #inpaint_tab' :
'#txt2img_tab'
);
for (const button of [...pasteButtons, ...pngButtons]) {
button.addEventListener('click', () => {
// The paste/send img generation info feature goes
// though gradio, which is pretty slow. Ideally we should
// observe the event when gradio has done the job, but
// that is not an easy task.
// Here we just do a 2 second delay until the refresh.
setTimeout(() => {
this.updateActiveState();
this.updateActiveUnitCount();
}, 2000);
});
}
}
/**
* Observer that triggers when the ControlNetUnit's accordion(tab) closes.
*/
attachAccordionStateObserver() {
new MutationObserver((mutationsList) => {
for(const mutation of mutationsList) {
if (mutation.type === 'attributes' && mutation.attributeName === 'class') {
const newState = mutation.target.classList.contains('open');
if (this.tabOpen != newState) {
this.tabOpen = newState;
if (newState) {
this.onAccordionOpen();
} else {
this.onAccordionClose();
}
}
}
}
}).observe(this.tab.querySelector('.label-wrap'), { attributes: true, attributeFilter: ['class'] });
}
onAccordionOpen() {
this.updateInputImageThumbnail();
}
onAccordionClose() {
this.updateInputImageThumbnail();
}
}
gradioApp().querySelectorAll('#controlnet').forEach(accordion => {
if (cnetAllAccordions.has(accordion)) return;
const tabs = [...accordion.querySelectorAll('.input-accordion')]
.map(tab => new ControlNetUnitTab(tab, accordion));
// On open of main extension accordion, if no unit is enabled,
// open unit 0 for edit.
const labelWrap = accordion.querySelector('.label-wrap');
const observerAccordionOpen = new MutationObserver(function (mutations) {
for (const mutation of mutations) {
if (mutation.target.classList.contains('open') &&
tabs.every(tab => !tab.enabledCheckbox.checked &&
!tab.tab.querySelector('.label-wrap').classList.contains('open'))
) {
tabs[0].tab.querySelector('.label-wrap').click();
}
}
});
observerAccordionOpen.observe(labelWrap, { attributes: true, attributeFilter: ['class'] });
cnetAllAccordions.add(accordion);
});
});
})();

17
extensions-builtin/sd_forge_controlnet/javascript/canvas.js Executable file
View File

@@ -0,0 +1,17 @@
(function () {
var hasApplied = false;
onUiUpdate(function () {
if (!hasApplied) {
if (typeof window.applyZoomAndPanIntegration === "function") {
hasApplied = true;
window.applyZoomAndPanIntegration("#txt2img_controlnet", Array.from({ length: 20 }, (_, i) => `#txt2img_controlnet_ControlNet-${i}_input_image`));
window.applyZoomAndPanIntegration("#img2img_controlnet", Array.from({ length: 20 }, (_, i) => `#img2img_controlnet_ControlNet-${i}_input_image`));
window.applyZoomAndPanIntegration("#txt2img_controlnet", ["#txt2img_controlnet_ControlNet_input_image"]);
window.applyZoomAndPanIntegration("#img2img_controlnet", ["#img2img_controlnet_ControlNet_input_image"]);
//console.log("window.applyZoomAndPanIntegration applied.");
} else {
//console.log("window.applyZoomAndPanIntegration is not available.");
}
}
});
})();

33
extensions-builtin/sd_forge_controlnet/javascript/modal.js Executable file
View File

@@ -0,0 +1,33 @@
(function () {
const cnetModalRegisteredElements = new Set();
onUiUpdate(() => {
// Get all the buttons that open a modal
const btns = gradioApp().querySelectorAll(".cnet-modal-open");
// Get all the <span> elements that close a modal
const spans = document.querySelectorAll(".cnet-modal-close");
// For each button, add a click event listener that opens the corresponding modal
btns.forEach((btn) => {
if (cnetModalRegisteredElements.has(btn)) return;
cnetModalRegisteredElements.add(btn);
const modalId = btn.id.replace('cnet-modal-open-', '');
const modal = document.getElementById("cnet-modal-" + modalId);
btn.addEventListener('click', () => {
modal.style.display = "block";
});
});
// For each <span> element, add a click event listener that closes the corresponding modal
spans.forEach((span) => {
if (cnetModalRegisteredElements.has(span)) return;
cnetModalRegisteredElements.add(span);
const modal = span.parentNode;
span.addEventListener('click', () => {
modal.style.display = "none";
});
});
});
})();

152
extensions-builtin/sd_forge_controlnet/javascript/openpose_editor.js Executable file
View File

@@ -0,0 +1,152 @@
(function () {
async function checkEditorAvailable() {
const LOCAL_EDITOR_PATH = '/openpose_editor_index';
const REMOTE_EDITOR_PATH = 'https://huchenlei.github.io/sd-webui-openpose-editor/';
async function testEditorPath(path) {
const res = await fetch(path);
return res.status === 200 ? path : null;
}
// Use local editor if the user has the extension installed. Fallback
// onto remote editor if the local editor is not ready yet.
// See https://github.com/huchenlei/sd-webui-openpose-editor/issues/53
// for more details.
return await testEditorPath(LOCAL_EDITOR_PATH) || await testEditorPath(REMOTE_EDITOR_PATH);
}
const cnetOpenposeEditorRegisteredElements = new Set();
let editorURL = null;
function loadOpenposeEditor() {
// Simulate an `input` DOM event for Gradio Textbox component. Needed after you edit its contents in javascript, otherwise your edits
// will only visible on web page and not sent to python.
function updateInput(target) {
let e = new Event("input", { bubbles: true })
Object.defineProperty(e, "target", { value: target })
target.dispatchEvent(e);
}
function navigateIframe(iframe, editorURL) {
function getPathname(rawURL) {
try {
return new URL(rawURL).pathname;
} catch (e) {
return rawURL;
}
}
return new Promise((resolve) => {
const darkThemeParam = document.body.classList.contains('dark') ?
new URLSearchParams({ theme: 'dark' }).toString() :
'';
window.addEventListener('message', (event) => {
const message = event.data;
if (message['ready']) resolve();
}, { once: true });
if ((editorURL.startsWith("http") ? iframe.src : getPathname(iframe.src)) !== editorURL) {
iframe.src = `${editorURL}?${darkThemeParam}`;
// By default assume 5 second is enough for the openpose editor
// to load.
setTimeout(resolve, 5000);
} else {
// If no navigation is required, immediately return.
resolve();
}
});
}
const tabs = gradioApp().querySelectorAll('#controlnet .input-accordion');
tabs.forEach(tab => {
if (cnetOpenposeEditorRegisteredElements.has(tab)) return;
cnetOpenposeEditorRegisteredElements.add(tab);
const generatedImageGroup = tab.querySelector('.cnet-generated-image-group');
const editButton = generatedImageGroup.querySelector('.cnet-edit-pose');
editButton.addEventListener('click', async () => {
const inputImageGroup = tab.querySelector('.cnet-input-image-group');
const inputImage = inputImageGroup.querySelector('.cnet-image img');
const downloadLink = generatedImageGroup.querySelector('.cnet-download-pose a');
const modalId = editButton.id.replace('cnet-modal-open-', '');
const modalIframe = generatedImageGroup.querySelector('.cnet-modal iframe');
if (!editorURL) {
editorURL = await checkEditorAvailable();
if (!editorURL) {
alert("No openpose editor available.")
}
}
await navigateIframe(modalIframe, editorURL);
modalIframe.contentWindow.postMessage({
modalId,
imageURL: inputImage ? inputImage.src : undefined,
poseURL: downloadLink.href,
}, '*');
// Focus the iframe so that the focus is no longer on the `Edit` button.
// Pressing space when the focus is on `Edit` button will trigger
// the click again to resend the frame message.
modalIframe.contentWindow.focus();
});
/*
* Writes the pose data URL to an link element on input image group.
* Click a hidden button to trigger a backend rendering of the pose JSON.
*
* The backend should:
* - Set the rendered pose image as preprocessor generated image.
*/
function updatePreviewPose(poseURL) {
const downloadLink = generatedImageGroup.querySelector('.cnet-download-pose a');
const renderButton = generatedImageGroup.querySelector('.cnet-render-pose');
const poseTextbox = generatedImageGroup.querySelector('.cnet-pose-json textarea');
const allowPreviewCheckbox = tab.querySelector('.cnet-allow-preview input');
if (!allowPreviewCheckbox.checked)
allowPreviewCheckbox.click();
// Only set href when download link exists and needs an update. `downloadLink`
// can be null when user closes preview and click `Upload JSON` button again.
// https://github.com/Mikubill/sd-webui-controlnet/issues/2308
if (downloadLink !== null)
downloadLink.href = poseURL;
poseTextbox.value = poseURL;
updateInput(poseTextbox);
renderButton.click();
}
// Updates preview image when edit is done.
window.addEventListener('message', (event) => {
const message = event.data;
const modalId = editButton.id.replace('cnet-modal-open-', '');
if (message.modalId !== modalId) return;
updatePreviewPose(message.poseURL);
const closeModalButton = generatedImageGroup.querySelector('.cnet-modal .cnet-modal-close');
closeModalButton.click();
});
const inputImageGroup = tab.querySelector('.cnet-input-image-group');
const uploadButton = inputImageGroup.querySelector('.cnet-upload-pose input');
// Updates preview image when JSON file is uploaded.
uploadButton.addEventListener('change', (event) => {
const file = event.target.files[0];
if (!file)
return;
const reader = new FileReader();
reader.onload = function (e) {
const contents = e.target.result;
const poseURL = `data:application/json;base64,${btoa(contents)}`;
updatePreviewPose(poseURL);
};
reader.readAsText(file);
// Reset the file input value so that uploading the same file still triggers callback.
event.target.value = '';
});
});
}
onUiUpdate(loadOpenposeEditor);
})();

435
extensions-builtin/sd_forge_controlnet/javascript/photopea.js Executable file
View File

@@ -0,0 +1,435 @@
(function () {
/*
MIT LICENSE
Copyright 2011 Jon Leighton
Permission is hereby granted, free of charge, to any person obtaining a copy of this software and
associated documentation files (the "Software"), to deal in the Software without restriction,
including without limitation the rights to use, copy, modify, merge, publish, distribute,
sublicense, and/or sell copies of the Software, and to permit persons to whom the Software is
furnished to do so, subject to the following conditions:
The above copyright notice and this permission notice shall be included in all copies or substantial
portions of the Software.
THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, FITNESS FOR A PARTICULAR
PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY
CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.
*/
// From: https://gist.github.com/jonleighton/958841
function base64ArrayBuffer(arrayBuffer) {
var base64 = ''
var encodings = 'ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz0123456789+/'
var bytes = new Uint8Array(arrayBuffer)
var byteLength = bytes.byteLength
var byteRemainder = byteLength % 3
var mainLength = byteLength - byteRemainder
var a, b, c, d
var chunk
// Main loop deals with bytes in chunks of 3
for (var i = 0; i < mainLength; i = i + 3) {
// Combine the three bytes into a single integer
chunk = (bytes[i] << 16) | (bytes[i + 1] << 8) | bytes[i + 2]
// Use bitmasks to extract 6-bit segments from the triplet
a = (chunk & 16515072) >> 18 // 16515072 = (2^6 - 1) << 18
b = (chunk & 258048) >> 12 // 258048 = (2^6 - 1) << 12
c = (chunk & 4032) >> 6 // 4032 = (2^6 - 1) << 6
d = chunk & 63 // 63 = 2^6 - 1
// Convert the raw binary segments to the appropriate ASCII encoding
base64 += encodings[a] + encodings[b] + encodings[c] + encodings[d]
}
// Deal with the remaining bytes and padding
if (byteRemainder == 1) {
chunk = bytes[mainLength]
a = (chunk & 252) >> 2 // 252 = (2^6 - 1) << 2
// Set the 4 least significant bits to zero
b = (chunk & 3) << 4 // 3 = 2^2 - 1
base64 += encodings[a] + encodings[b] + '=='
} else if (byteRemainder == 2) {
chunk = (bytes[mainLength] << 8) | bytes[mainLength + 1]
a = (chunk & 64512) >> 10 // 64512 = (2^6 - 1) << 10
b = (chunk & 1008) >> 4 // 1008 = (2^6 - 1) << 4
// Set the 2 least significant bits to zero
c = (chunk & 15) << 2 // 15 = 2^4 - 1
base64 += encodings[a] + encodings[b] + encodings[c] + '='
}
return base64
}
// Turn a base64 string into a blob.
// From https://gist.github.com/gauravmehla/7a7dfd87dd7d1b13697b6e894426615f
function b64toBlob(b64Data, contentType, sliceSize) {
var contentType = contentType || '';
var sliceSize = sliceSize || 512;
var byteCharacters = atob(b64Data);
var byteArrays = [];
for (var offset = 0; offset < byteCharacters.length; offset += sliceSize) {
var slice = byteCharacters.slice(offset, offset + sliceSize);
var byteNumbers = new Array(slice.length);
for (var i = 0; i < slice.length; i++) {
byteNumbers[i] = slice.charCodeAt(i);
}
var byteArray = new Uint8Array(byteNumbers);
byteArrays.push(byteArray);
}
return new Blob(byteArrays, { type: contentType });
}
function createBlackImageBase64(width, height) {
// Create a canvas element
var canvas = document.createElement('canvas');
canvas.width = width;
canvas.height = height;
// Get the context of the canvas
var ctx = canvas.getContext('2d');
// Fill the canvas with black color
ctx.fillStyle = 'black';
ctx.fillRect(0, 0, width, height);
// Get the base64 encoded string
var base64Image = canvas.toDataURL('image/png');
return base64Image;
}
// Functions to be called within photopea context.
// Start of photopea functions
function pasteImage(base64image) {
app.open(base64image, null, /* asSmart */ true);
app.echoToOE("success");
}
function setLayerNames(names) {
const layers = app.activeDocument.layers;
if (layers.length !== names.length) {
console.error("layer length does not match names length");
echoToOE("error");
return;
}
for (let i = 0; i < names.length; i++) {
const layer = layers[i];
layer.name = names[i];
}
app.echoToOE("success");
}
function removeLayersWithNames(names) {
const layers = app.activeDocument.layers;
for (let i = 0; i < layers.length; i++) {
const layer = layers[i];
if (names.includes(layer.name)) {
layer.remove();
}
}
app.echoToOE("success");
}
function getAllLayerNames() {
const layers = app.activeDocument.layers;
const names = [];
for (let i = 0; i < layers.length; i++) {
const layer = layers[i];
names.push(layer.name);
}
app.echoToOE(JSON.stringify(names));
}
// Hides all layers except the current one, outputs the whole image, then restores the previous
// layers state.
function exportSelectedLayerOnly(format, layerName) {
// Gets all layers recursively, including the ones inside folders.
function getAllArtLayers(document) {
let allArtLayers = [];
for (let i = 0; i < document.layers.length; i++) {
const currentLayer = document.layers[i];
allArtLayers.push(currentLayer);
if (currentLayer.typename === "LayerSet") {
allArtLayers = allArtLayers.concat(getAllArtLayers(currentLayer));
}
}
return allArtLayers;
}
function makeLayerVisible(layer) {
let currentLayer = layer;
while (currentLayer != app.activeDocument) {
currentLayer.visible = true;
if (currentLayer.parent.typename != 'Document') {
currentLayer = currentLayer.parent;
} else {
break;
}
}
}
const allLayers = getAllArtLayers(app.activeDocument);
// Make all layers except the currently selected one invisible, and store
// their initial state.
const layerStates = [];
for (let i = 0; i < allLayers.length; i++) {
const layer = allLayers[i];
layerStates.push(layer.visible);
}
// Hide all layers to begin with
for (let i = 0; i < allLayers.length; i++) {
const layer = allLayers[i];
layer.visible = false;
}
for (let i = 0; i < allLayers.length; i++) {
const layer = allLayers[i];
const selected = layer.name === layerName;
if (selected) {
makeLayerVisible(layer);
}
}
app.activeDocument.saveToOE(format);
for (let i = 0; i < allLayers.length; i++) {
const layer = allLayers[i];
layer.visible = layerStates[i];
}
}
function hasActiveDocument() {
app.echoToOE(app.documents.length > 0 ? "true" : "false");
}
// End of photopea functions
const MESSAGE_END_ACK = "done";
const MESSAGE_ERROR = "error";
const PHOTOPEA_URL = "https://www.photopea.com/";
class PhotopeaContext {
constructor(photopeaIframe) {
this.photopeaIframe = photopeaIframe;
this.timeout = 1000;
}
navigateIframe() {
const iframe = this.photopeaIframe;
const editorURL = PHOTOPEA_URL;
return new Promise(async (resolve) => {
if (iframe.src !== editorURL) {
iframe.src = editorURL;
// Stop waiting after 10s.
setTimeout(resolve, 10000);
// Testing whether photopea is able to accept message.
while (true) {
try {
await this.invoke(hasActiveDocument);
break;
} catch (e) {
console.log("Keep waiting for photopea to accept message.");
}
}
this.timeout = 5000; // Restore to a longer timeout in normal messaging.
}
resolve();
});
}
// From https://github.com/huchenlei/stable-diffusion-ps-pea/blob/main/src/Photopea.ts
postMessageToPhotopea(message) {
return new Promise((resolve, reject) => {
const responseDataPieces = [];
let hasError = false;
const photopeaMessageHandle = (event) => {
if (event.source !== this.photopeaIframe.contentWindow) {
return;
}
// Filter out the ping messages
if (typeof event.data === 'string' && event.data.includes('MSFAPI#')) {
return;
}
// Ignore "done" when no data has been received. The "done" can come from
// MSFAPI ping.
if (event.data === MESSAGE_END_ACK && responseDataPieces.length === 0) {
return;
}
if (event.data === MESSAGE_END_ACK) {
window.removeEventListener("message", photopeaMessageHandle);
if (hasError) {
reject('Photopea Error.');
} else {
resolve(responseDataPieces.length === 1 ? responseDataPieces[0] : responseDataPieces);
}
} else if (event.data === MESSAGE_ERROR) {
responseDataPieces.push(event.data);
hasError = true;
} else {
responseDataPieces.push(event.data);
}
};
window.addEventListener("message", photopeaMessageHandle);
setTimeout(() => reject("Photopea message timeout"), this.timeout);
this.photopeaIframe.contentWindow.postMessage(message, "*");
});
}
// From https://github.com/huchenlei/stable-diffusion-ps-pea/blob/main/src/Photopea.ts
async invoke(func, ...args) {
await this.navigateIframe();
const message = `${func.toString()} ${func.name}(${args.map(arg => JSON.stringify(arg)).join(',')});`;
try {
return await this.postMessageToPhotopea(message);
} catch (e) {
throw `Failed to invoke ${func.name}. ${e}.`;
}
}
/**
* Fetch detected maps from each ControlNet units.
* Create a new photopea document.
* Add those detected maps to the created document.
*/
async fetchFromControlNet(tabs) {
if (tabs.length === 0) return;
const isImg2Img = tabs[0].querySelector('.cnet-mask-upload').id.includes('img2img');
const generationType = isImg2Img ? 'img2img' : 'txt2img';
const width = gradioApp().querySelector(`#${generationType}_width input[type=number]`).value;
const height = gradioApp().querySelector(`#${generationType}_height input[type=number]`).value;
const layerNames = ["background"];
await this.invoke(pasteImage, createBlackImageBase64(width, height));
await new Promise(r => setTimeout(r, 200));
for (const [i, tab] of tabs.entries()) {
const generatedImage = tab.querySelector('.cnet-generated-image-group .cnet-image img');
if (!generatedImage) continue;
await this.invoke(pasteImage, generatedImage.src);
// Wait 200ms for pasting to fully complete so that we do not ended up with 2 separate
// documents.
await new Promise(r => setTimeout(r, 200));
layerNames.push(`unit-${i}`);
}
await this.invoke(removeLayersWithNames, layerNames);
await this.invoke(setLayerNames, layerNames.reverse());
}
/**
* Send the images in the active photopea document back to each ControlNet units.
*/
async sendToControlNet(tabs) {
// Gradio's image widgets are inputs. To set the image in one, we set the image on the input and
// force it to refresh.
function setImageOnInput(imageInput, file) {
// Createa a data transfer element to set as the data in the input.
const dt = new DataTransfer();
dt.items.add(file);
const list = dt.files;
// Actually set the image in the image widget.
imageInput.files = list;
// Foce the image widget to update with the new image, after setting its source files.
const event = new Event('change', {
'bubbles': true,
"composed": true
});
imageInput.dispatchEvent(event);
}
function sendToControlNetUnit(b64Image, index) {
const tab = tabs[index];
// Upload image to output image element.
const outputImage = tab.querySelector('.cnet-photopea-output');
const outputImageUpload = outputImage.querySelector('input[type="file"]');
setImageOnInput(outputImageUpload, new File([b64toBlob(b64Image, "image/png")], "photopea_output.png"));
// Make sure `UsePreviewAsInput` checkbox is checked.
const checkbox = tab.querySelector('.cnet-preview-as-input input[type="checkbox"]');
if (!checkbox.checked) {
checkbox.click();
}
}
const layerNames =
JSON.parse(await this.invoke(getAllLayerNames))
.filter(name => /unit-\d+/.test(name));
for (const layerName of layerNames) {
const arrayBuffer = await this.invoke(exportSelectedLayerOnly, 'PNG', layerName);
const b64Image = base64ArrayBuffer(arrayBuffer);
const layerIndex = Number.parseInt(layerName.split('-')[1]);
sendToControlNetUnit(b64Image, layerIndex);
}
}
}
let photopeaWarningShown = false;
function firstTimeUserPrompt() {
if (opts.controlnet_photopea_warning){
const photopeaPopupMsg = "you are about to connect to https://photopea.com\n" +
"- Click OK: proceed.\n" +
"- Click Cancel: abort.\n" +
"Photopea integration can be disabled in Settings > ControlNet > Disable photopea edit.\n" +
"This popup can be disabled in Settings > ControlNet > Photopea popup warning.";
if (photopeaWarningShown || confirm(photopeaPopupMsg)) photopeaWarningShown = true;
else return false;
}
return true;
}
const cnetRegisteredAccordions = new Set();
function loadPhotopea() {
function registerCallbacks(accordion) {
const photopeaMainTrigger = accordion.querySelector('.cnet-photopea-main-trigger');
// Photopea edit feature disabled.
if (!photopeaMainTrigger) {
console.log("ControlNet photopea edit disabled.");
return;
}
const closeModalButton = accordion.querySelector('.cnet-photopea-edit .cnet-modal-close');
const tabs = accordion.querySelectorAll('.controlnet .input-accordion');
const photopeaIframe = accordion.querySelector('.photopea-iframe');
const photopeaContext = new PhotopeaContext(photopeaIframe, tabs);
tabs.forEach(tab => {
const photopeaChildTrigger = tab.querySelector('.cnet-photopea-child-trigger');
photopeaChildTrigger.addEventListener('click', async () => {
if (!firstTimeUserPrompt()) return;
photopeaMainTrigger.click();
if (await photopeaContext.invoke(hasActiveDocument) === "false") {
await photopeaContext.fetchFromControlNet(tabs);
}
});
});
accordion.querySelector('.photopea-fetch').addEventListener('click', () => photopeaContext.fetchFromControlNet(tabs));
accordion.querySelector('.photopea-send').addEventListener('click', () => {
photopeaContext.sendToControlNet(tabs)
closeModalButton.click();
});
}
const accordions = gradioApp().querySelectorAll('#controlnet');
accordions.forEach(accordion => {
if (cnetRegisteredAccordions.has(accordion)) return;
registerCallbacks(accordion);
cnetRegisteredAccordions.add(accordion);
});
}
onUiUpdate(loadPhotopea);
})();

138
extensions-builtin/sd_forge_controlnet/lib_controlnet/api.py Executable file
View File

@@ -0,0 +1,138 @@
from typing import List
import numpy as np
from fastapi import FastAPI, Body
from fastapi.exceptions import HTTPException
from PIL import Image
import gradio as gr
from modules.api import api
from .global_state import (
get_all_preprocessor_names,
get_all_controlnet_names,
get_preprocessor,
get_all_preprocessor_tags,
select_control_type,
)
from .utils import judge_image_type
from .logging import logger
def encode_to_base64(image):
if isinstance(image, str):
return image
elif not judge_image_type(image):
return "Detect result is not image"
elif isinstance(image, Image.Image):
return api.encode_pil_to_base64(image)
elif isinstance(image, np.ndarray):
return encode_np_to_base64(image)
else:
logger.warn("Unable to encode image.")
return ""
def encode_np_to_base64(image):
pil = Image.fromarray(image)
return api.encode_pil_to_base64(pil)
def controlnet_api(_: gr.Blocks, app: FastAPI):
@app.get("/controlnet/model_list")
async def model_list():
up_to_date_model_list = get_all_controlnet_names()
logger.debug(up_to_date_model_list)
return {"model_list": up_to_date_model_list}
@app.get("/controlnet/module_list")
async def module_list():
module_list = get_all_preprocessor_names()
logger.debug(module_list)
return {
"module_list": module_list,
# TODO: Add back module detail.
# "module_detail": external_code.get_modules_detail(alias_names),
}
@app.get("/controlnet/control_types")
async def control_types():
def format_control_type(
filtered_preprocessor_list,
filtered_model_list,
default_option,
default_model,
):
control_dict = {
"module_list": filtered_preprocessor_list,
"model_list": filtered_model_list,
"default_option": default_option,
"default_model": default_model,
}
return control_dict
return {
"control_types": {
control_type: format_control_type(*select_control_type(control_type))
for control_type in get_all_preprocessor_tags()
}
}
@app.post("/controlnet/detect")
async def detect(
controlnet_module: str = Body("none", title="Controlnet Module"),
controlnet_input_images: List[str] = Body([], title="Controlnet Input Images"),
controlnet_processor_res: int = Body(
512, title="Controlnet Processor Resolution"
),
controlnet_threshold_a: float = Body(64, title="Controlnet Threshold a"),
controlnet_threshold_b: float = Body(64, title="Controlnet Threshold b"),
):
processor_module = get_preprocessor(controlnet_module)
if processor_module is None:
raise HTTPException(status_code=422, detail="Module not available")
if len(controlnet_input_images) == 0:
raise HTTPException(status_code=422, detail="No image selected")
logger.debug(
f"Detecting {str(len(controlnet_input_images))} images with the {controlnet_module} module."
)
results = []
poses = []
for input_image in controlnet_input_images:
img = np.array(api.decode_base64_to_image(input_image)).astype('uint8')
class JsonAcceptor:
def __init__(self) -> None:
self.value = None
def accept(self, json_dict: dict) -> None:
self.value = json_dict
json_acceptor = JsonAcceptor()
results.append(
processor_module(
img,
resolution=controlnet_processor_res,
slider_1=controlnet_threshold_a,
slider_2=controlnet_threshold_b,
json_pose_callback=json_acceptor.accept,
)
)
if "openpose" in controlnet_module:
assert json_acceptor.value is not None
poses.append(json_acceptor.value)
results64 = [encode_to_base64(img) for img in results]
res = {"images": results64, "info": "Success"}
if poses:
res["poses"] = poses
return res

1113
extensions-builtin/sd_forge_controlnet/lib_controlnet/controlnet_ui/controlnet_ui_group.py Executable file
View File

File diff suppressed because it is too large Load Diff

38
extensions-builtin/sd_forge_controlnet/lib_controlnet/controlnet_ui/modal.py Executable file
View File

@@ -0,0 +1,38 @@
import gradio as gr
from typing import List
class ModalInterface(gr.Interface):
modal_id_counter = 0
def __init__(
self,
html_content: str,
open_button_text: str,
open_button_classes: List[str] = [],
open_button_extra_attrs: str = ''
):
self.html_content = html_content
self.open_button_text = open_button_text
self.open_button_classes = open_button_classes
self.open_button_extra_attrs = open_button_extra_attrs
self.modal_id = ModalInterface.modal_id_counter
ModalInterface.modal_id_counter += 1
def __call__(self):
return self.create_modal()
def create_modal(self, visible=True):
html_code = f"""
<div id="cnet-modal-{self.modal_id}" class="cnet-modal">
<span class="cnet-modal-close">&times;</span>
<div class="cnet-modal-content">
{self.html_content}
</div>
</div>
<div id="cnet-modal-open-{self.modal_id}"
class="cnet-modal-open {' '.join(self.open_button_classes)}"
{self.open_button_extra_attrs}
>{self.open_button_text}</div>
"""
return gr.HTML(value=html_code, visible=visible)

154
extensions-builtin/sd_forge_controlnet/lib_controlnet/controlnet_ui/openpose_editor.py Executable file
View File

@@ -0,0 +1,154 @@
import base64
import gradio as gr
import json
from typing import List, Dict, Any, Tuple
from annotator.openpose import decode_json_as_poses, draw_poses
from annotator.openpose.animalpose import draw_animalposes
from lib_controlnet.controlnet_ui.modal import ModalInterface
from modules import shared
from lib_controlnet.logging import logger
def parse_data_url(data_url: str):
# Split the URL at the comma
media_type, data = data_url.split(",", 1)
# Check if the data is base64-encoded
assert ";base64" in media_type
# Decode the base64 data
return base64.b64decode(data)
def encode_data_url(json_string: str) -> str:
base64_encoded_json = base64.b64encode(json_string.encode("utf-8")).decode("utf-8")
return f"data:application/json;base64,{base64_encoded_json}"
class OpenposeEditor(object):
# Filename used when user click the download link.
download_file = "pose.json"
# URL the openpose editor is mounted on.
editor_url = "/openpose_editor_index"
def __init__(self) -> None:
self.render_button = None
self.pose_input = None
self.download_link = None
self.upload_link = None
self.modal = None
def render_edit(self):
"""Renders the buttons in preview image control button group."""
# The hidden button to trigger a re-render of generated image.
self.render_button = gr.Button(visible=False, elem_classes=["cnet-render-pose"])
# The hidden element that stores the pose json for backend retrieval.
# The front-end javascript will write the edited JSON data to the element.
self.pose_input = gr.Textbox(visible=False, elem_classes=["cnet-pose-json"])
self.modal = ModalInterface(
# Use about:blank here as placeholder so that the iframe does not
# immediately navigate. Most of controlnet units do not need
# openpose editor active. Only navigate when the user first click
# 'Edit'. The navigation logic is in `openpose_editor.js`.
f'<iframe src="about:blank"></iframe>',
open_button_text="Edit",
open_button_classes=["cnet-edit-pose"],
open_button_extra_attrs=f'title="Send pose to {OpenposeEditor.editor_url} for edit."',
).create_modal(visible=False)
self.download_link = gr.HTML(
value=f"""<a href='' download='{OpenposeEditor.download_file}'>JSON</a>""",
visible=False,
elem_classes=["cnet-download-pose"],
)
def render_upload(self):
"""Renders the button in input image control button group."""
self.upload_link = gr.HTML(
value="""
<label>Upload JSON</label>
<input type="file" accept=".json"/>
""",
visible=False,
elem_classes=["cnet-upload-pose"],
)
def register_callbacks(
self,
generated_image: gr.Image,
use_preview_as_input: gr.Checkbox,
model: gr.Dropdown,
):
def render_pose(pose_url: str) -> Tuple[Dict, Dict]:
json_string = parse_data_url(pose_url).decode("utf-8")
poses, animals, height, width = decode_json_as_poses(
json.loads(json_string)
)
logger.info("Preview as input is enabled.")
return (
# Generated image.
gr.update(
value=(
draw_poses(
poses,
height,
width,
draw_body=True,
draw_hand=True,
draw_face=True,
)
if poses
else draw_animalposes(animals, height, width)
),
visible=True,
),
# Use preview as input.
gr.update(value=True),
# Self content.
*self.update(json_string),
)
self.render_button.click(
fn=render_pose,
inputs=[self.pose_input],
outputs=[generated_image.background, use_preview_as_input, *self.outputs()],
)
def update_upload_link(model: str) -> Dict:
return gr.update(visible="openpose" in model.lower())
model.change(fn=update_upload_link, inputs=[model], outputs=[self.upload_link])
def outputs(self) -> List[Any]:
return [
self.download_link,
self.modal,
]
def update(self, json_string: str) -> List[Dict]:
"""
Called when there is a new JSON pose value generated by running
preprocessor.
Args:
json_string: The new JSON string generated by preprocessor.
Returns:
An gr.update event.
"""
hint = "Download the pose as .json file"
html = f"""<a href='{encode_data_url(json_string)}'
download='{OpenposeEditor.download_file}' title="{hint}">
JSON</a>"""
visible = json_string != ""
return [
# Download link update.
gr.update(value=html, visible=visible),
# Modal update.
gr.update(
visible=visible
and not shared.opts.data.get("controlnet_disable_openpose_edit", False)
),
]

182
extensions-builtin/sd_forge_controlnet/lib_controlnet/controlnet_ui/photopea.py Executable file
View File

@@ -0,0 +1,182 @@
import gradio as gr
from lib_controlnet.controlnet_ui.modal import ModalInterface
PHOTOPEA_LOGO = """
<svg version="1.1" id="Layer_1" xmlns="http://www.w3.org/2000/svg" xmlns:xlink="http://www.w3.org/1999/xlink" x="0px" y="0px"
width="100%" viewBox="0 0 256 256" enable-background="new 0 0 256 256" xml:space="preserve"
style="width: 0.75rem; height 0.75rem; margin-left: 2px;"
>
<path fill="#18A497" opacity="1.000000" stroke="none"
d="
M1.000000,228.000000
C1.000000,162.312439 1.000000,96.624878 1.331771,30.719650
C2.026278,30.171114 2.594676,29.904894 2.721949,29.500008
C6.913495,16.165672 15.629609,7.322631 28.880219,2.875538
C29.404272,2.699659 29.633436,1.645129 30.000000,1.000000
C95.687561,1.000000 161.375122,1.000000 227.258057,1.317018
C227.660217,1.893988 227.815079,2.296565 228.081207,2.393433
C241.304657,7.206383 250.980164,15.550970 255.215851,29.410040
C255.321625,29.756128 256.383850,29.809898 257.000000,30.000000
C257.000000,95.687561 257.000000,161.375122 256.682983,227.257858
C256.106049,227.659790 255.699371,227.815521 255.607178,228.080658
C250.953033,241.462830 242.292618,250.822968 228.591782,255.214935
C228.239929,255.327698 228.190491,256.383820 228.000000,257.000000
C175.312439,257.000000 122.624878,257.000000 69.468582,256.531342
C68.672188,244.948196 68.218323,233.835587 68.052299,222.718674
C67.885620,211.557587 67.886772,200.390717 68.027298,189.229050
C68.255180,171.129044 68.084618,152.997421 69.151917,134.942368
C70.148468,118.083969 77.974228,103.689308 89.758743,91.961365
C104.435837,77.354736 122.313736,69.841736 143.417328,69.901505
C168.662338,69.972984 186.981964,90.486633 187.961487,114.156334
C189.042435,140.277435 166.783981,163.607941 140.303482,160.823074
C137.092346,160.485382 133.490692,158.365784 131.192612,155.987366
C126.434669,151.063141 126.975357,144.720825 129.168777,138.834930
C131.533630,132.489014 137.260605,130.548050 143.413757,130.046677
C150.288467,129.486496 156.424942,123.757378 157.035324,117.320816
C157.953949,107.633820 150.959381,101.769096 145.533951,101.194389
C132.238846,99.786079 120.699944,104.963120 111.676735,114.167313
C102.105782,123.930222 97.469498,136.194061 99.003151,150.234955
C100.540352,164.308228 107.108505,175.507980 118.864334,183.311539
C128.454544,189.677597 138.866959,191.786957 150.657837,190.245651
C166.242554,188.208420 179.874283,182.443329 191.251801,172.056793
C209.355011,155.530380 217.848694,134.938721 216.116119,110.085892
C214.834335,91.699440 207.721039,76.015915 195.289444,62.978828
C175.658447,42.391735 150.833389,37.257801 123.833740,42.281937
C98.675804,46.963364 78.315033,60.084667 62.208153,80.157814
C46.645889,99.552216 39.305275,121.796379 39.149052,146.201981
C38.912663,183.131317 39.666767,220.067017 40.000000,257.000000
C36.969406,257.000000 33.938812,257.000000 30.705070,256.668213
C30.298622,256.078369 30.144913,255.669220 29.884926,255.583878
C16.317770,251.131058 7.127485,242.317780 2.778462,228.591797
C2.667588,228.241821 1.613958,228.190567 1.000000,228.000000
z"/>
<path fill="#000000" opacity="1.000000" stroke="none"
d="
M40.468658,257.000000
C39.666767,220.067017 38.912663,183.131317 39.149052,146.201981
C39.305275,121.796379 46.645889,99.552216 62.208153,80.157814
C78.315033,60.084667 98.675804,46.963364 123.833740,42.281937
C150.833389,37.257801 175.658447,42.391735 195.289444,62.978828
C207.721039,76.015915 214.834335,91.699440 216.116119,110.085892
C217.848694,134.938721 209.355011,155.530380 191.251801,172.056793
C179.874283,182.443329 166.242554,188.208420 150.657837,190.245651
C138.866959,191.786957 128.454544,189.677597 118.864334,183.311539
C107.108505,175.507980 100.540352,164.308228 99.003151,150.234955
C97.469498,136.194061 102.105782,123.930222 111.676735,114.167313
C120.699944,104.963120 132.238846,99.786079 145.533951,101.194389
C150.959381,101.769096 157.953949,107.633820 157.035324,117.320816
C156.424942,123.757378 150.288467,129.486496 143.413757,130.046677
C137.260605,130.548050 131.533630,132.489014 129.168777,138.834930
C126.975357,144.720825 126.434669,151.063141 131.192612,155.987366
C133.490692,158.365784 137.092346,160.485382 140.303482,160.823074
C166.783981,163.607941 189.042435,140.277435 187.961487,114.156334
C186.981964,90.486633 168.662338,69.972984 143.417328,69.901505
C122.313736,69.841736 104.435837,77.354736 89.758743,91.961365
C77.974228,103.689308 70.148468,118.083969 69.151917,134.942368
C68.084618,152.997421 68.255180,171.129044 68.027298,189.229050
C67.886772,200.390717 67.885620,211.557587 68.052299,222.718674
C68.218323,233.835587 68.672188,244.948196 68.999924,256.531342
C59.645771,257.000000 50.291542,257.000000 40.468658,257.000000
z"/>
<path fill="#000000" opacity="1.000000" stroke="none"
d="
M257.000000,29.531342
C256.383850,29.809898 255.321625,29.756128 255.215851,29.410040
C250.980164,15.550970 241.304657,7.206383 228.081207,2.393433
C227.815079,2.296565 227.660217,1.893988 227.726715,1.317018
C237.593155,1.000000 247.186295,1.000000 257.000000,1.000000
C257.000000,10.353075 257.000000,19.707878 257.000000,29.531342
z"/>
<path fill="#000000" opacity="1.000000" stroke="none"
d="
M228.468658,257.000000
C228.190491,256.383820 228.239929,255.327698 228.591782,255.214935
C242.292618,250.822968 250.953033,241.462830 255.607178,228.080658
C255.699371,227.815521 256.106049,227.659790 256.682983,227.726517
C257.000000,237.593155 257.000000,247.186295 257.000000,257.000000
C247.646927,257.000000 238.292114,257.000000 228.468658,257.000000
z"/>
<path fill="#000000" opacity="1.000000" stroke="none"
d="
M1.000000,228.468658
C1.613958,228.190567 2.667588,228.241821 2.778462,228.591797
C7.127485,242.317780 16.317770,251.131058 29.884926,255.583878
C30.144913,255.669220 30.298622,256.078369 30.250959,256.668213
C20.406853,257.000000 10.813705,257.000000 1.000000,257.000000
C1.000000,247.646927 1.000000,238.292114 1.000000,228.468658
z"/>
<path fill="#000000" opacity="1.000000" stroke="none"
d="
M29.531342,1.000000
C29.633436,1.645129 29.404272,2.699659 28.880219,2.875538
C15.629609,7.322631 6.913495,16.165672 2.721949,29.500008
C2.594676,29.904894 2.026278,30.171114 1.331771,30.250992
C1.000000,20.406855 1.000000,10.813709 1.000000,1.000000
C10.353074,1.000000 19.707878,1.000000 29.531342,1.000000
z"/>
</svg>"""
class Photopea(object):
def __init__(self) -> None:
self.modal = None
self.triggers = []
self.render_editor()
def render_editor(self):
"""Render the editor modal."""
with gr.Group(elem_classes=["cnet-photopea-edit"]):
self.modal = ModalInterface(
# Use about:blank here as placeholder so that the iframe does not
# immediately navigate. Only navigate when the user first click
# 'Edit'. The navigation logic is in `photopea.js`.
f"""
<div class="photopea-button-group">
<button class="photopea-button photopea-fetch">Fetch from ControlNet</button>
<button class="photopea-button photopea-send">Send to ControlNet</button>
</div>
<iframe class="photopea-iframe" src="about:blank"></iframe>
""",
open_button_text="Edit",
open_button_classes=["cnet-photopea-main-trigger"],
open_button_extra_attrs="hidden",
).create_modal(visible=True)
def render_child_trigger(self):
self.triggers.append(
gr.HTML(
f"""<div class="cnet-photopea-child-trigger">
Edit {PHOTOPEA_LOGO}
</div>"""
)
)
def attach_photopea_output(self, generated_image: gr.Image):
"""Called in ControlNetUiGroup to attach preprocessor preview image Gradio element
as the photopea output. If the front-end directly change the img HTML element's src
to reflect the edited image result from photopea, the backend won't be notified.
In this method we let the front-end upload the result image an invisible gr.Image
instance and mirrors the value to preprocessor preview gr.Image. This is because
the generated image gr.Image instance is inferred to be an output image by Gradio
and has no ability to accept image upload directly.
Arguments:
generated_image: preprocessor result Gradio Image output element.
Returns:
None
"""
output = gr.Image(
visible=False,
source="upload",
type="numpy",
elem_classes=[f"cnet-photopea-output"],
)
output.upload(
fn=lambda img: img,
inputs=[output],
outputs=[generated_image],
)

39
extensions-builtin/sd_forge_controlnet/lib_controlnet/enums.py Executable file
View File

@@ -0,0 +1,39 @@
from enum import Enum
class HiResFixOption(Enum):
BOTH = "Both"
LOW_RES_ONLY = "Low res only"
HIGH_RES_ONLY = "High res only"
@staticmethod
def from_value(value) -> "HiResFixOption":
if isinstance(value, str) and value.startswith("HiResFixOption."):
_, field = value.split(".")
return getattr(HiResFixOption, field)
if isinstance(value, str):
return HiResFixOption(value)
elif isinstance(value, int):
return [x for x in HiResFixOption][value]
else:
assert isinstance(value, HiResFixOption)
return value
@property
def low_res_enabled(self) -> bool:
return self in (HiResFixOption.BOTH, HiResFixOption.LOW_RES_ONLY)
@property
def high_res_enabled(self) -> bool:
return self in (HiResFixOption.BOTH, HiResFixOption.HIGH_RES_ONLY)
class InputMode(Enum):
# Single image to a single ControlNet unit.
SIMPLE = "simple"
# Input is a directory. N generations. Each generation takes 1 input image
# from the directory.
BATCH = "batch"
# Input is a directory. 1 generation. Each generation takes N input image
# from the directory.
MERGE = "merge"

238
extensions-builtin/sd_forge_controlnet/lib_controlnet/external_code.py Executable file
View File

@@ -0,0 +1,238 @@
from dataclasses import dataclass
from enum import Enum
from typing import List, Optional, Union, Dict, TypedDict
import numpy as np
from modules import shared
from lib_controlnet.logging import logger
from lib_controlnet.enums import InputMode, HiResFixOption
from modules.api import api
def get_api_version() -> int:
return 2
class ControlMode(Enum):
"""
The improved guess mode.
"""
BALANCED = "Balanced"
PROMPT = "My prompt is more important"
CONTROL = "ControlNet is more important"
class BatchOption(Enum):
DEFAULT = "All ControlNet units for all images in a batch"
SEPARATE = "Each ControlNet unit for each image in a batch"
class ResizeMode(Enum):
"""
Resize modes for ControlNet input images.
"""
RESIZE = "Just Resize"
INNER_FIT = "Crop and Resize"
OUTER_FIT = "Resize and Fill"
def int_value(self):
if self == ResizeMode.RESIZE:
return 0
elif self == ResizeMode.INNER_FIT:
return 1
elif self == ResizeMode.OUTER_FIT:
return 2
assert False, "NOTREACHED"
resize_mode_aliases = {
'Inner Fit (Scale to Fit)': 'Crop and Resize',
'Outer Fit (Shrink to Fit)': 'Resize and Fill',
'Scale to Fit (Inner Fit)': 'Crop and Resize',
'Envelope (Outer Fit)': 'Resize and Fill',
}
def resize_mode_from_value(value: Union[str, int, ResizeMode]) -> ResizeMode:
if isinstance(value, str):
return ResizeMode(resize_mode_aliases.get(value, value))
elif isinstance(value, int):
assert value >= 0
if value == 3: # 'Just Resize (Latent upscale)'
return ResizeMode.RESIZE
if value >= len(ResizeMode):
logger.warning(f'Unrecognized ResizeMode int value {value}. Fall back to RESIZE.')
return ResizeMode.RESIZE
return [e for e in ResizeMode][value]
else:
return value
def control_mode_from_value(value: Union[str, int, ControlMode]) -> ControlMode:
if isinstance(value, str):
return ControlMode(value)
elif isinstance(value, int):
return [e for e in ControlMode][value]
else:
return value
def visualize_inpaint_mask(img):
if img.ndim == 3 and img.shape[2] == 4:
result = img.copy()
mask = result[:, :, 3]
mask = 255 - mask // 2
result[:, :, 3] = mask
return np.ascontiguousarray(result.copy())
return img
def pixel_perfect_resolution(
image: np.ndarray,
target_H: int,
target_W: int,
resize_mode: ResizeMode,
) -> int:
"""
Calculate the estimated resolution for resizing an image while preserving aspect ratio.
The function first calculates scaling factors for height and width of the image based on the target
height and width. Then, based on the chosen resize mode, it either takes the smaller or the larger
scaling factor to estimate the new resolution.
If the resize mode is OUTER_FIT, the function uses the smaller scaling factor, ensuring the whole image
fits within the target dimensions, potentially leaving some empty space.
If the resize mode is not OUTER_FIT, the function uses the larger scaling factor, ensuring the target
dimensions are fully filled, potentially cropping the image.
After calculating the estimated resolution, the function prints some debugging information.
Args:
image (np.ndarray): A 3D numpy array representing an image. The dimensions represent [height, width, channels].
target_H (int): The target height for the image.
target_W (int): The target width for the image.
resize_mode (ResizeMode): The mode for resizing.
Returns:
int: The estimated resolution after resizing.
"""
raw_H, raw_W, _ = image.shape
k0 = float(target_H) / float(raw_H)
k1 = float(target_W) / float(raw_W)
if resize_mode == ResizeMode.OUTER_FIT:
estimation = min(k0, k1) * float(min(raw_H, raw_W))
else:
estimation = max(k0, k1) * float(min(raw_H, raw_W))
logger.debug(f"Pixel Perfect Computation:")
logger.debug(f"resize_mode = {resize_mode}")
logger.debug(f"raw_H = {raw_H}")
logger.debug(f"raw_W = {raw_W}")
logger.debug(f"target_H = {target_H}")
logger.debug(f"target_W = {target_W}")
logger.debug(f"estimation = {estimation}")
return int(np.round(estimation))
class GradioImageMaskPair(TypedDict):
"""Represents the dict object from Gradio's image component if `tool="sketch"`
is specified.
{
"image": np.ndarray,
"mask": np.ndarray,
}
"""
image: np.ndarray
mask: np.ndarray
@dataclass
class ControlNetUnit:
input_mode: InputMode = InputMode.SIMPLE
use_preview_as_input: bool = False
batch_image_dir: str = ''
batch_mask_dir: str = ''
batch_input_gallery: Optional[List[str]] = None
batch_mask_gallery: Optional[List[str]] = None
generated_image: Optional[np.ndarray] = None
mask_image: Optional[GradioImageMaskPair] = None
mask_image_fg: Optional[GradioImageMaskPair] = None
hr_option: Union[HiResFixOption, int, str] = HiResFixOption.BOTH
enabled: bool = True
module: str = "None"
model: str = "None"
weight: float = 1.0
image: Optional[GradioImageMaskPair] = None
image_fg: Optional[GradioImageMaskPair] = None
resize_mode: Union[ResizeMode, int, str] = ResizeMode.INNER_FIT
processor_res: int = -1
threshold_a: float = -1
threshold_b: float = -1
guidance_start: float = 0.0
guidance_end: float = 1.0
pixel_perfect: bool = False
control_mode: Union[ControlMode, int, str] = ControlMode.BALANCED
save_detected_map: bool = True
@staticmethod
def infotext_fields():
"""Fields that should be included in infotext.
You should define a Gradio element with exact same name in ControlNetUiGroup
as well, so that infotext can wire the value to correct field when pasting
infotext.
"""
return (
"module",
"model",
"weight",
"resize_mode",
"processor_res",
"threshold_a",
"threshold_b",
"guidance_start",
"guidance_end",
"pixel_perfect",
"control_mode",
"hr_option",
)
@staticmethod
def from_dict(d: Dict) -> "ControlNetUnit":
"""Create ControlNetUnit from dict. This is primarily used to convert
API json dict to ControlNetUnit."""
unit = ControlNetUnit(
**{k: v for k, v in d.items() if k in vars(ControlNetUnit)}
)
if isinstance(unit.image, str):
unit.image = np.array(api.decode_base64_to_image(unit.image)).astype('uint8')
if isinstance(unit.mask_image, str):
unit.mask_image = np.array(api.decode_base64_to_image(unit.mask_image)).astype('uint8')
return unit
# Backward Compatible
UiControlNetUnit = ControlNetUnit
def to_base64_nparray(encoding: str):
"""
Convert a base64 image into the image type the extension uses
"""
return np.array(api.decode_base64_to_image(encoding)).astype('uint8')
def get_max_models_num():
"""
Fetch the maximum number of allowed ControlNet models.
"""
max_models_num = shared.opts.data.get("control_net_unit_count", 3)
return max_models_num

171
extensions-builtin/sd_forge_controlnet/lib_controlnet/global_state.py Executable file
View File

@@ -0,0 +1,171 @@
import os.path
import stat
from collections import OrderedDict
from modules import shared, sd_models
from modules_forge.shared import controlnet_dir, supported_preprocessors
from typing import Dict, Tuple, List
CN_MODEL_EXTS = [".pt", ".pth", ".ckpt", ".safetensors", ".bin", ".patch"]
def traverse_all_files(curr_path, model_list):
f_list = [
(os.path.join(curr_path, entry.name), entry.stat())
for entry in os.scandir(curr_path)
if os.path.isdir(curr_path)
]
for f_info in f_list:
fname, fstat = f_info
if os.path.splitext(fname)[1] in CN_MODEL_EXTS:
model_list.append(f_info)
elif stat.S_ISDIR(fstat.st_mode):
model_list = traverse_all_files(fname, model_list)
return model_list
def get_all_models(sort_by, filter_by, path):
res = OrderedDict()
fileinfos = traverse_all_files(path, [])
filter_by = filter_by.strip(" ")
if len(filter_by) != 0:
fileinfos = [x for x in fileinfos if filter_by.lower()
in os.path.basename(x[0]).lower()]
if sort_by == "name":
fileinfos = sorted(fileinfos, key=lambda x: os.path.basename(x[0]))
elif sort_by == "date":
fileinfos = sorted(fileinfos, key=lambda x: -x[1].st_mtime)
elif sort_by == "path name":
fileinfos = sorted(fileinfos)
for finfo in fileinfos:
filename = finfo[0]
name = os.path.splitext(os.path.basename(filename))[0]
# Prevent a hypothetical "None.pt" from being listed.
if name != "None":
res[name + f" [{sd_models.model_hash(filename)}]"] = filename
return res
controlnet_filename_dict = {'None': 'model.safetensors'}
controlnet_names = ['None']
def get_preprocessor(name):
return supported_preprocessors.get(name, None)
def get_default_preprocessor(tag):
ps = get_filtered_preprocessor_names(tag)
assert len(ps) > 0
return ps[0] if len(ps) == 1 else ps[1]
def get_sorted_preprocessors():
preprocessors = [p for k, p in supported_preprocessors.items() if k != 'None']
preprocessors = sorted(preprocessors, key=lambda x: str(x.sorting_priority).zfill(8) + x.name)[::-1]
results = OrderedDict()
results['None'] = supported_preprocessors['None']
for p in preprocessors:
results[p.name] = p
return results
def get_all_controlnet_names():
return controlnet_names
def get_controlnet_filename(controlnet_name):
return controlnet_filename_dict[controlnet_name]
def get_all_preprocessor_names():
return list(get_sorted_preprocessors().keys())
def get_all_preprocessor_tags():
tags = []
for k, p in supported_preprocessors.items():
tags += p.tags
tags = list(set(tags))
tags = sorted(tags)
return ['All'] + tags
def get_filtered_preprocessors(tag):
if tag == 'All':
return supported_preprocessors
return {k: v for k, v in get_sorted_preprocessors().items() if tag in v.tags or k == 'None'}
def get_filtered_preprocessor_names(tag):
return list(get_filtered_preprocessors(tag).keys())
def get_filtered_controlnet_names(tag):
filtered_preprocessors = get_filtered_preprocessors(tag)
model_filename_filters = []
for p in filtered_preprocessors.values():
model_filename_filters += p.model_filename_filters
return [x for x in controlnet_names if x == 'None' or any(f.lower() in x.lower() for f in model_filename_filters)]
def update_controlnet_filenames():
global controlnet_filename_dict, controlnet_names
controlnet_filename_dict = {'None': 'model.safetensors'}
controlnet_names = ['None']
ext_dirs = (shared.opts.data.get("control_net_models_path", None), getattr(shared.cmd_opts, 'controlnet_dir', None))
extra_lora_paths = (extra_lora_path for extra_lora_path in ext_dirs
if extra_lora_path is not None and os.path.exists(extra_lora_path))
paths = [controlnet_dir, *extra_lora_paths]
for path in paths:
sort_by = shared.opts.data.get("control_net_models_sort_models_by", "name")
filter_by = shared.opts.data.get("control_net_models_name_filter", "")
found = get_all_models(sort_by, filter_by, path)
controlnet_filename_dict.update(found)
controlnet_names = list(controlnet_filename_dict.keys())
return
def select_control_type(
control_type: str,
) -> Tuple[List[str], List[str], str, str]:
global controlnet_names
pattern = control_type.lower()
all_models = list(controlnet_names)
if pattern == "all":
preprocessors = get_sorted_preprocessors().values()
return [
[p.name for p in preprocessors],
all_models,
'none', # default option
"None" # default model
]
filtered_model_list = get_filtered_controlnet_names(control_type)
if pattern == "none":
filtered_model_list.append("None")
assert len(filtered_model_list) > 0, "'None' model should always be available."
if len(filtered_model_list) == 1:
default_model = "None"
else:
default_model = filtered_model_list[1]
for x in filtered_model_list:
if "11" in x.split("[")[0]:
default_model = x
break
return (
get_filtered_preprocessor_names(control_type),
filtered_model_list,
get_default_preprocessor(control_type),
default_model
)

121
extensions-builtin/sd_forge_controlnet/lib_controlnet/infotext.py Executable file
View File

@@ -0,0 +1,121 @@
from typing import List, Tuple, Union
import gradio as gr
from modules.processing import StableDiffusionProcessing
from lib_controlnet import external_code
from lib_controlnet.logging import logger
def field_to_displaytext(fieldname: str) -> str:
return " ".join([word.capitalize() for word in fieldname.split("_")])
def displaytext_to_field(text: str) -> str:
return "_".join([word.lower() for word in text.split(" ")])
def parse_value(value: str) -> Union[str, float, int, bool]:
if value in ("True", "False"):
return value == "True"
try:
return int(value)
except ValueError:
try:
return float(value)
except ValueError:
return value # Plain string.
def serialize_unit(unit: external_code.ControlNetUnit) -> str:
log_value = {
field_to_displaytext(field): getattr(unit, field)
for field in external_code.ControlNetUnit.infotext_fields()
if getattr(unit, field) != -1
# Note: exclude hidden slider values.
}
if not all("," not in str(v) and ":" not in str(v) for v in log_value.values()):
logger.error(f"Unexpected tokens encountered:\n{log_value}")
return ""
return ", ".join(f"{field}: {value}" for field, value in log_value.items())
def parse_unit(text: str) -> external_code.ControlNetUnit:
return external_code.ControlNetUnit(
enabled=True,
**{
displaytext_to_field(key): parse_value(value)
for item in text.split(",")
for (key, value) in (item.strip().split(": "),)
},
)
class Infotext(object):
def __init__(self) -> None:
self.infotext_fields: List[Tuple[gr.components.IOComponent, str]] = []
self.paste_field_names: List[str] = []
@staticmethod
def unit_prefix(unit_index: int) -> str:
return f"ControlNet {unit_index}"
def register_unit(self, unit_index: int, uigroup) -> None:
"""Register the unit's UI group. By regsitering the unit, A1111 will be
able to paste values from infotext to IOComponents.
Args:
unit_index: The index of the ControlNet unit
uigroup: The ControlNetUiGroup instance that contains all gradio
iocomponents.
"""
unit_prefix = Infotext.unit_prefix(unit_index)
for field in external_code.ControlNetUnit.infotext_fields():
# Every field in ControlNetUnit should have a corresponding
# IOComponent in ControlNetUiGroup.
io_component = getattr(uigroup, field)
component_locator = f"{unit_prefix} {field}"
self.infotext_fields.append((io_component, component_locator))
self.paste_field_names.append(component_locator)
@staticmethod
def write_infotext(
units: List[external_code.ControlNetUnit], p: StableDiffusionProcessing
):
"""Write infotext to `p`."""
p.extra_generation_params.update(
{
Infotext.unit_prefix(i): serialize_unit(unit)
for i, unit in enumerate(units)
if unit.enabled
}
)
@staticmethod
def on_infotext_pasted(infotext: str, results: dict) -> None:
"""Parse ControlNet infotext string and write result to `results` dict."""
updates = {}
for k, v in results.items():
if not k.startswith("ControlNet"):
continue
assert isinstance(v, str), f"Expect string but got {v}."
try:
for field, value in vars(parse_unit(v)).items():
if field == "image":
continue
if value is None:
logger.debug(f"InfoText: Skipping {field} because value is None.")
continue
component_locator = f"{k} {field}"
updates[component_locator] = value
logger.debug(f"InfoText: Setting {component_locator} = {value}")
except Exception as e:
logger.warn(
f"Failed to parse infotext, legacy format infotext is no longer supported:\n{v}\n{e}"
)
results.update(updates)

41
extensions-builtin/sd_forge_controlnet/lib_controlnet/logging.py Executable file
View File

@@ -0,0 +1,41 @@
import logging
import copy
import sys
from modules import shared
class ColoredFormatter(logging.Formatter):
COLORS = {
"DEBUG": "\033[0;36m", # CYAN
"INFO": "\033[0;32m", # GREEN
"WARNING": "\033[0;33m", # YELLOW
"ERROR": "\033[0;31m", # RED
"CRITICAL": "\033[0;37;41m", # WHITE ON RED
"RESET": "\033[0m", # RESET COLOR
}
def format(self, record):
colored_record = copy.copy(record)
levelname = colored_record.levelname
seq = self.COLORS.get(levelname, self.COLORS["RESET"])
colored_record.levelname = f"{seq}{levelname}{self.COLORS['RESET']}"
return super().format(colored_record)
# Create a new logger
logger = logging.getLogger("ControlNet")
logger.propagate = False
# Add handler if we don't have one.
if not logger.handlers:
handler = logging.StreamHandler(sys.stdout)
handler.setFormatter(
ColoredFormatter("%(asctime)s - %(name)s - %(levelname)s - %(message)s")
)
logger.addHandler(handler)
# Configure logger
loglevel_string = getattr(shared.cmd_opts, "controlnet_loglevel", "INFO")
loglevel = getattr(logging, loglevel_string.upper(), None)
logger.setLevel(loglevel)

88
extensions-builtin/sd_forge_controlnet/lib_controlnet/lvminthin.py Executable file
View File

@@ -0,0 +1,88 @@
# High Quality Edge Thinning using Pure Python
# Written by Lvmin Zhang
# 2023 April
# Stanford University
# If you use this, please Cite "High Quality Edge Thinning using Pure Python", Lvmin Zhang, In Mikubill/sd-webui-controlnet.
import cv2
import numpy as np
lvmin_kernels_raw = [
np.array([
[-1, -1, -1],
[0, 1, 0],
[1, 1, 1]
], dtype=np.int32),
np.array([
[0, -1, -1],
[1, 1, -1],
[0, 1, 0]
], dtype=np.int32)
]
lvmin_kernels = []
lvmin_kernels += [np.rot90(x, k=0, axes=(0, 1)) for x in lvmin_kernels_raw]
lvmin_kernels += [np.rot90(x, k=1, axes=(0, 1)) for x in lvmin_kernels_raw]
lvmin_kernels += [np.rot90(x, k=2, axes=(0, 1)) for x in lvmin_kernels_raw]
lvmin_kernels += [np.rot90(x, k=3, axes=(0, 1)) for x in lvmin_kernels_raw]
lvmin_prunings_raw = [
np.array([
[-1, -1, -1],
[-1, 1, -1],
[0, 0, -1]
], dtype=np.int32),
np.array([
[-1, -1, -1],
[-1, 1, -1],
[-1, 0, 0]
], dtype=np.int32)
]
lvmin_prunings = []
lvmin_prunings += [np.rot90(x, k=0, axes=(0, 1)) for x in lvmin_prunings_raw]
lvmin_prunings += [np.rot90(x, k=1, axes=(0, 1)) for x in lvmin_prunings_raw]
lvmin_prunings += [np.rot90(x, k=2, axes=(0, 1)) for x in lvmin_prunings_raw]
lvmin_prunings += [np.rot90(x, k=3, axes=(0, 1)) for x in lvmin_prunings_raw]
def remove_pattern(x, kernel):
objects = cv2.morphologyEx(x, cv2.MORPH_HITMISS, kernel)
objects = np.where(objects > 127)
x[objects] = 0
return x, objects[0].shape[0] > 0
def thin_one_time(x, kernels):
y = x
is_done = True
for k in kernels:
y, has_update = remove_pattern(y, k)
if has_update:
is_done = False
return y, is_done
def lvmin_thin(x, prunings=True):
y = x
for i in range(32):
y, is_done = thin_one_time(y, lvmin_kernels)
if is_done:
break
if prunings:
y, _ = thin_one_time(y, lvmin_prunings)
return y
def nake_nms(x):
f1 = np.array([[0, 0, 0], [1, 1, 1], [0, 0, 0]], dtype=np.uint8)
f2 = np.array([[0, 1, 0], [0, 1, 0], [0, 1, 0]], dtype=np.uint8)
f3 = np.array([[1, 0, 0], [0, 1, 0], [0, 0, 1]], dtype=np.uint8)
f4 = np.array([[0, 0, 1], [0, 1, 0], [1, 0, 0]], dtype=np.uint8)
y = np.zeros_like(x)
for f in [f1, f2, f3, f4]:
np.putmask(y, cv2.dilate(x, kernel=f) == x, x)
return y

362
extensions-builtin/sd_forge_controlnet/lib_controlnet/utils.py Executable file
View File

@@ -0,0 +1,362 @@
from typing import Optional
from modules import processing
from lib_controlnet import external_code
from modules_forge.utils import HWC3
from PIL import Image, ImageFilter, ImageOps
from lib_controlnet.lvminthin import lvmin_thin, nake_nms
import torch
import os
import functools
import time
import base64
import numpy as np
import safetensors.torch
import cv2
import logging
from typing import Any, Callable, Dict, List
from lib_controlnet.logging import logger
def load_state_dict(ckpt_path, location="cpu"):
_, extension = os.path.splitext(ckpt_path)
if extension.lower() == ".safetensors":
state_dict = safetensors.torch.load_file(ckpt_path, device=location)
else:
state_dict = torch.load(ckpt_path, map_location=torch.device(location))
state_dict = get_state_dict(state_dict)
logger.info(f"Loaded state_dict from [{ckpt_path}]")
return state_dict
def get_state_dict(d):
return d.get("state_dict", d)
def ndarray_lru_cache(max_size: int = 128, typed: bool = False):
"""
Decorator to enable caching for functions with numpy array arguments.
Numpy arrays are mutable, and thus not directly usable as hash keys.
The idea here is to wrap the incoming arguments with type `np.ndarray`
as `HashableNpArray` so that `lru_cache` can correctly handles `np.ndarray`
arguments.
`HashableNpArray` functions exactly the same way as `np.ndarray` except
having `__hash__` and `__eq__` overriden.
"""
def decorator(func: Callable):
"""The actual decorator that accept function as input."""
class HashableNpArray(np.ndarray):
def __new__(cls, input_array):
# Input array is an instance of ndarray.
# The view makes the input array and returned array share the same data.
obj = np.asarray(input_array).view(cls)
return obj
def __eq__(self, other) -> bool:
return np.array_equal(self, other)
def __hash__(self):
# Hash the bytes representing the data of the array.
return hash(self.tobytes())
@functools.lru_cache(maxsize=max_size, typed=typed)
def cached_func(*args, **kwargs):
"""This function only accepts `HashableNpArray` as input params."""
return func(*args, **kwargs)
# Preserves original function.__name__ and __doc__.
@functools.wraps(func)
def decorated_func(*args, **kwargs):
"""The decorated function that delegates the original function."""
def convert_item(item: Any):
if isinstance(item, np.ndarray):
return HashableNpArray(item)
if isinstance(item, tuple):
return tuple(convert_item(i) for i in item)
return item
args = [convert_item(arg) for arg in args]
kwargs = {k: convert_item(arg) for k, arg in kwargs.items()}
return cached_func(*args, **kwargs)
return decorated_func
return decorator
def timer_decorator(func):
"""Time the decorated function and output the result to debug logger."""
if logger.level != logging.DEBUG:
return func
@functools.wraps(func)
def wrapper(*args, **kwargs):
start_time = time.time()
result = func(*args, **kwargs)
end_time = time.time()
duration = end_time - start_time
# Only report function that are significant enough.
if duration > 1e-3:
logger.debug(f"{func.__name__} ran in: {duration:.3f} sec")
return result
return wrapper
class TimeMeta(type):
""" Metaclass to record execution time on all methods of the
child class. """
def __new__(cls, name, bases, attrs):
for attr_name, attr_value in attrs.items():
if callable(attr_value):
attrs[attr_name] = timer_decorator(attr_value)
return super().__new__(cls, name, bases, attrs)
# svgsupports
svgsupport = False
try:
import io
from svglib.svglib import svg2rlg
from reportlab.graphics import renderPM
svgsupport = True
except ImportError:
pass
def svg_preprocess(inputs: Dict, preprocess: Callable):
if not inputs:
return None
if inputs["image"].startswith("data:image/svg+xml;base64,") and svgsupport:
svg_data = base64.b64decode(
inputs["image"].replace("data:image/svg+xml;base64,", "")
)
drawing = svg2rlg(io.BytesIO(svg_data))
png_data = renderPM.drawToString(drawing, fmt="PNG")
encoded_string = base64.b64encode(png_data)
base64_str = str(encoded_string, "utf-8")
base64_str = "data:image/png;base64," + base64_str
inputs["image"] = base64_str
return preprocess(inputs)
def get_unique_axis0(data):
arr = np.asanyarray(data)
idxs = np.lexsort(arr.T)
arr = arr[idxs]
unique_idxs = np.empty(len(arr), dtype=np.bool_)
unique_idxs[:1] = True
unique_idxs[1:] = np.any(arr[:-1, :] != arr[1:, :], axis=-1)
return arr[unique_idxs]
def read_image(img_path: str) -> str:
"""Read image from specified path and return a base64 string."""
img = cv2.imread(img_path)
_, bytes = cv2.imencode(".png", img)
encoded_image = base64.b64encode(bytes).decode("utf-8")
return encoded_image
def read_image_dir(img_dir: str, suffixes=('.png', '.jpg', '.jpeg', '.webp')) -> List[str]:
"""Try read all images in given img_dir."""
images = []
for filename in os.listdir(img_dir):
if filename.endswith(suffixes):
img_path = os.path.join(img_dir, filename)
try:
images.append(read_image(img_path))
except IOError:
logger.error(f"Error opening {img_path}")
return images
def align_dim_latent(x: int) -> int:
""" Align the pixel dimension (w/h) to latent dimension.
Stable diffusion 1:8 ratio for latent/pixel, i.e.,
1 latent unit == 8 pixel unit."""
return (x // 8) * 8
def prepare_mask(
mask: Image.Image, p: processing.StableDiffusionProcessing
) -> Image.Image:
"""
Prepare an image mask for the inpainting process.
This function takes as input a PIL Image object and an instance of the
StableDiffusionProcessing class, and performs the following steps to prepare the mask:
1. Convert the mask to grayscale (mode "L").
2. If the 'inpainting_mask_invert' attribute of the processing instance is True,
invert the mask colors.
3. If the 'mask_blur' attribute of the processing instance is greater than 0,
apply a Gaussian blur to the mask with a radius equal to 'mask_blur'.
Args:
mask (Image.Image): The input mask as a PIL Image object.
p (processing.StableDiffusionProcessing): An instance of the StableDiffusionProcessing class
containing the processing parameters.
Returns:
mask (Image.Image): The prepared mask as a PIL Image object.
"""
mask = mask.convert("L")
if getattr(p, "inpainting_mask_invert", False):
mask = ImageOps.invert(mask)
if hasattr(p, 'mask_blur_x'):
if getattr(p, "mask_blur_x", 0) > 0:
np_mask = np.array(mask)
kernel_size = 2 * int(2.5 * p.mask_blur_x + 0.5) + 1
np_mask = cv2.GaussianBlur(np_mask, (kernel_size, 1), p.mask_blur_x)
mask = Image.fromarray(np_mask)
if getattr(p, "mask_blur_y", 0) > 0:
np_mask = np.array(mask)
kernel_size = 2 * int(2.5 * p.mask_blur_y + 0.5) + 1
np_mask = cv2.GaussianBlur(np_mask, (1, kernel_size), p.mask_blur_y)
mask = Image.fromarray(np_mask)
else:
if getattr(p, "mask_blur", 0) > 0:
mask = mask.filter(ImageFilter.GaussianBlur(p.mask_blur))
return mask
def set_numpy_seed(p: processing.StableDiffusionProcessing) -> Optional[int]:
"""
Set the random seed for NumPy based on the provided parameters.
Args:
p (processing.StableDiffusionProcessing): The instance of the StableDiffusionProcessing class.
Returns:
Optional[int]: The computed random seed if successful, or None if an exception occurs.
This function sets the random seed for NumPy using the seed and subseed values from the given instance of
StableDiffusionProcessing. If either seed or subseed is -1, it uses the first value from `all_seeds`.
Otherwise, it takes the maximum of the provided seed value and 0.
The final random seed is computed by adding the seed and subseed values, applying a bitwise AND operation
with 0xFFFFFFFF to ensure it fits within a 32-bit integer.
"""
try:
tmp_seed = int(p.all_seeds[0] if p.seed == -1 else max(int(p.seed), 0))
tmp_subseed = int(p.all_seeds[0] if p.subseed == -1 else max(int(p.subseed), 0))
seed = (tmp_seed + tmp_subseed) & 0xFFFFFFFF
np.random.seed(seed)
return seed
except Exception as e:
logger.warning(e)
logger.warning('Warning: Failed to use consistent random seed.')
return None
def safe_numpy(x):
# A very safe method to make sure that Apple/Mac works
y = x
# below is very boring but do not change these. If you change these Apple or Mac may fail.
y = y.copy()
y = np.ascontiguousarray(y)
y = y.copy()
return y
def high_quality_resize(x, size):
# Written by lvmin
# Super high-quality control map up-scaling, considering binary, seg, and one-pixel edges
if x.shape[0] != size[1] or x.shape[1] != size[0]:
new_size_is_smaller = (size[0] * size[1]) < (x.shape[0] * x.shape[1])
new_size_is_bigger = (size[0] * size[1]) > (x.shape[0] * x.shape[1])
unique_color_count = len(get_unique_axis0(x.reshape(-1, x.shape[2])))
is_one_pixel_edge = False
is_binary = False
if unique_color_count == 2:
is_binary = np.min(x) < 16 and np.max(x) > 240
if is_binary:
xc = x
xc = cv2.erode(xc, np.ones(shape=(3, 3), dtype=np.uint8), iterations=1)
xc = cv2.dilate(xc, np.ones(shape=(3, 3), dtype=np.uint8), iterations=1)
one_pixel_edge_count = np.where(xc < x)[0].shape[0]
all_edge_count = np.where(x > 127)[0].shape[0]
is_one_pixel_edge = one_pixel_edge_count * 2 > all_edge_count
if 2 < unique_color_count < 200:
interpolation = cv2.INTER_NEAREST
elif new_size_is_smaller:
interpolation = cv2.INTER_AREA
else:
interpolation = cv2.INTER_CUBIC # Must be CUBIC because we now use nms. NEVER CHANGE THIS
y = cv2.resize(x, size, interpolation=interpolation)
if is_binary:
y = np.mean(y.astype(np.float32), axis=2).clip(0, 255).astype(np.uint8)
if is_one_pixel_edge:
y = nake_nms(y)
_, y = cv2.threshold(y, 0, 255, cv2.THRESH_BINARY + cv2.THRESH_OTSU)
y = lvmin_thin(y, prunings=new_size_is_bigger)
else:
_, y = cv2.threshold(y, 0, 255, cv2.THRESH_BINARY + cv2.THRESH_OTSU)
y = np.stack([y] * 3, axis=2)
else:
y = x
return y
def crop_and_resize_image(detected_map, resize_mode, h, w, fill_border_with_255=False):
if resize_mode == external_code.ResizeMode.RESIZE:
detected_map = high_quality_resize(detected_map, (w, h))
detected_map = safe_numpy(detected_map)
return detected_map
old_h, old_w, _ = detected_map.shape
old_w = float(old_w)
old_h = float(old_h)
k0 = float(h) / old_h
k1 = float(w) / old_w
safeint = lambda x: int(np.round(x))
if resize_mode == external_code.ResizeMode.OUTER_FIT:
k = min(k0, k1)
borders = np.concatenate([detected_map[0, :, :], detected_map[-1, :, :], detected_map[:, 0, :], detected_map[:, -1, :]], axis=0)
high_quality_border_color = np.median(borders, axis=0).astype(detected_map.dtype)
if fill_border_with_255:
high_quality_border_color = np.zeros_like(high_quality_border_color) + 255
high_quality_background = np.tile(high_quality_border_color[None, None], [h, w, 1])
detected_map = high_quality_resize(detected_map, (safeint(old_w * k), safeint(old_h * k)))
new_h, new_w, _ = detected_map.shape
pad_h = max(0, (h - new_h) // 2)
pad_w = max(0, (w - new_w) // 2)
high_quality_background[pad_h:pad_h + new_h, pad_w:pad_w + new_w] = detected_map
detected_map = high_quality_background
detected_map = safe_numpy(detected_map)
return detected_map
else:
k = max(k0, k1)
detected_map = high_quality_resize(detected_map, (safeint(old_w * k), safeint(old_h * k)))
new_h, new_w, _ = detected_map.shape
pad_h = max(0, (new_h - h) // 2)
pad_w = max(0, (new_w - w) // 2)
detected_map = detected_map[pad_h:pad_h+h, pad_w:pad_w+w]
detected_map = safe_numpy(detected_map)
return detected_map
def judge_image_type(img):
return isinstance(img, np.ndarray) and img.ndim == 3 and int(img.shape[2]) in [3, 4]

13
extensions-builtin/sd_forge_controlnet/preload.py Executable file
View File

@@ -0,0 +1,13 @@
def preload(parser):
parser.add_argument(
"--controlnet-loglevel",
default="INFO",
choices=["DEBUG", "INFO", "WARNING", "ERROR", "CRITICAL"],
help="Set the log level (DEBUG, INFO, WARNING, ERROR, CRITICAL)",
)
parser.add_argument(
"--controlnet-tracemalloc",
action="store_true",
help="Enable memory tracing.",
default=None,
)

5
extensions-builtin/sd_forge_controlnet/requirements.txt Executable file
View File

@@ -0,0 +1,5 @@
fvcore
mediapipe
onnxruntime
opencv-python>=4.8.0
svglib

617
extensions-builtin/sd_forge_controlnet/scripts/controlnet.py Executable file
View File

@@ -0,0 +1,617 @@
import os
from typing import Dict, Optional, Tuple, List, Union
import cv2
import torch
import modules.scripts as scripts
from modules import shared, script_callbacks, masking, images
from modules.ui_components import InputAccordion
from modules.api.api import decode_base64_to_image
import gradio as gr
from lib_controlnet import global_state, external_code
from lib_controlnet.external_code import ControlNetUnit
from lib_controlnet.utils import align_dim_latent, set_numpy_seed, crop_and_resize_image, \
prepare_mask, judge_image_type
from lib_controlnet.controlnet_ui.controlnet_ui_group import ControlNetUiGroup
from lib_controlnet.controlnet_ui.photopea import Photopea
from lib_controlnet.logging import logger
from modules.processing import StableDiffusionProcessingImg2Img, StableDiffusionProcessingTxt2Img, \
StableDiffusionProcessing
from lib_controlnet.infotext import Infotext
from modules_forge.utils import HWC3, numpy_to_pytorch
from lib_controlnet.enums import HiResFixOption
from lib_controlnet.api import controlnet_api
import numpy as np
import functools
from PIL import Image
from modules_forge.shared import try_load_supported_control_model
from modules_forge.supported_controlnet import ControlModelPatcher
# Gradio 3.32 bug fix
import tempfile
gradio_tempfile_path = os.path.join(tempfile.gettempdir(), 'gradio')
os.makedirs(gradio_tempfile_path, exist_ok=True)
global_state.update_controlnet_filenames()
@functools.lru_cache(maxsize=shared.opts.data.get("control_net_model_cache_size", 5))
def cached_controlnet_loader(filename):
return try_load_supported_control_model(filename)
class ControlNetCachedParameters:
def __init__(self):
self.preprocessor = None
self.model = None
self.control_cond = None
self.control_cond_for_hr_fix = None
self.control_mask = None
self.control_mask_for_hr_fix = None
class ControlNetForForgeOfficial(scripts.Script):
sorting_priority = 10
def title(self):
return "ControlNet"
def show(self, is_img2img):
return scripts.AlwaysVisible
def ui(self, is_img2img):
infotext = Infotext()
ui_groups = []
controls = []
max_models = shared.opts.data.get("control_net_unit_count", 3)
gen_type = "img2img" if is_img2img else "txt2img"
elem_id_tabname = gen_type + "_controlnet"
default_unit = ControlNetUnit(enabled=False, module="None", model="None")
with gr.Group(elem_id=elem_id_tabname):
with gr.Accordion(f"ControlNet Integrated", open=False, elem_id="controlnet",
elem_classes=["controlnet"]):
photopea = (
Photopea()
if not shared.opts.data.get("controlnet_disable_photopea_edit", False)
else None
)
with gr.Row(elem_id=elem_id_tabname + "_accordions", elem_classes="accordions"):
for i in range(max_models):
with InputAccordion(
value=False,
label=f"ControlNet Unit {i}",
elem_classes=["cnet-unit-enabled-accordion"], # Class on accordion
):
group = ControlNetUiGroup(is_img2img, default_unit, photopea)
ui_groups.append(group)
controls.append(group.render(f"ControlNet-{i}", elem_id_tabname))
for i, ui_group in enumerate(ui_groups):
infotext.register_unit(i, ui_group)
if shared.opts.data.get("control_net_sync_field_args", True):
self.infotext_fields = infotext.infotext_fields
self.paste_field_names = infotext.paste_field_names
return tuple(controls)
def get_enabled_units(self, units):
# Parse dict from API calls.
units = [
ControlNetUnit.from_dict(unit) if isinstance(unit, dict) else unit
for unit in units
]
assert all(isinstance(unit, ControlNetUnit) for unit in units)
enabled_units = [x for x in units if x.enabled]
return enabled_units
@staticmethod
def try_crop_image_with_a1111_mask(
p: StableDiffusionProcessing,
unit: ControlNetUnit,
input_image: np.ndarray,
resize_mode: external_code.ResizeMode,
preprocessor
) -> np.ndarray:
a1111_mask_image: Optional[Image.Image] = getattr(p, "image_mask", None)
is_only_masked_inpaint = (
issubclass(type(p), StableDiffusionProcessingImg2Img) and
p.inpaint_full_res and
a1111_mask_image is not None
)
if (
preprocessor.corp_image_with_a1111_mask_when_in_img2img_inpaint_tab
and is_only_masked_inpaint
):
logger.info("Crop input image based on A1111 mask.")
input_image = [input_image[:, :, i] for i in range(input_image.shape[2])]
input_image = [Image.fromarray(x) for x in input_image]
mask = prepare_mask(a1111_mask_image, p)
crop_region = masking.get_crop_region(np.array(mask), p.inpaint_full_res_padding)
crop_region = masking.expand_crop_region(crop_region, p.width, p.height, mask.width, mask.height)
input_image = [
images.resize_image(resize_mode.int_value(), i, mask.width, mask.height)
for i in input_image
]
input_image = [x.crop(crop_region) for x in input_image]
input_image = [
images.resize_image(external_code.ResizeMode.OUTER_FIT.int_value(), x, p.width, p.height)
for x in input_image
]
input_image = [np.asarray(x)[:, :, 0] for x in input_image]
input_image = np.stack(input_image, axis=2)
return input_image
def get_input_data(self, p, unit, preprocessor, h, w):
logger.info(f'ControlNet Input Mode: {unit.input_mode}')
image_list = []
resize_mode = external_code.resize_mode_from_value(unit.resize_mode)
if unit.input_mode == external_code.InputMode.MERGE:
for idx, item in enumerate(unit.batch_input_gallery):
img_path = item[0]
logger.info(f'Try to read image: {img_path}')
img = np.ascontiguousarray(cv2.imread(img_path)[:, :, ::-1]).copy()
mask = None
if unit.batch_mask_gallery is not None and len(unit.batch_mask_gallery) > 0:
if len(unit.batch_mask_gallery) >= len(unit.batch_input_gallery):
mask_path = unit.batch_mask_gallery[idx]['name']
else:
mask_path = unit.batch_mask_gallery[0]['name']
mask = np.ascontiguousarray(cv2.imread(mask_path)[:, :, ::-1]).copy()
if img is not None:
image_list.append([img, mask])
elif unit.input_mode == external_code.InputMode.BATCH:
image_list = []
image_extensions = ['.jpg', '.jpeg', '.png', '.bmp']
batch_image_files = shared.listfiles(unit.batch_image_dir)
for batch_modifier in getattr(unit, 'batch_modifiers', []):
batch_image_files = batch_modifier(batch_image_files, p)
for idx, filename in enumerate(batch_image_files):
if any(filename.lower().endswith(ext) for ext in image_extensions):
img_path = os.path.join(unit.batch_image_dir, filename)
logger.info(f'Try to read image: {img_path}')
img = np.ascontiguousarray(cv2.imread(img_path)[:, :, ::-1]).copy()
mask = None
if unit.batch_mask_dir:
batch_mask_files = shared.listfiles(unit.batch_mask_dir)
if len(batch_mask_files) >= len(batch_image_files):
mask_path = batch_mask_files[idx]
else:
mask_path = batch_mask_files[0]
mask_path = os.path.join(unit.batch_mask_dir, mask_path)
mask = np.ascontiguousarray(cv2.imread(mask_path)[:, :, ::-1]).copy()
if img is not None:
image_list.append([img, mask])
else:
a1111_i2i_image = getattr(p, "init_images", [None])[0]
a1111_i2i_mask = getattr(p, "image_mask", None)
using_a1111_data = False
unit_image = unit.image
unit_image_fg = unit.image_fg[:, :, 3] if unit.image_fg is not None else None
if unit.use_preview_as_input and unit.generated_image is not None:
image = unit.generated_image
elif unit.image is None:
resize_mode = external_code.resize_mode_from_value(p.resize_mode)
image = HWC3(np.asarray(a1111_i2i_image))
using_a1111_data = True
elif (unit_image < 5).all() and (unit_image_fg > 5).any():
image = unit_image_fg
else:
image = unit_image
if not isinstance(image, np.ndarray):
raise ValueError("controlnet is enabled but no input image is given")
image = HWC3(image)
unit_mask_image = unit.mask_image
unit_mask_image_fg = unit.mask_image_fg[:, :, 3] if unit.mask_image_fg is not None else None
if using_a1111_data:
mask = HWC3(np.asarray(a1111_i2i_mask)) if a1111_i2i_mask is not None else None
elif unit_mask_image_fg is not None and (unit_mask_image_fg > 5).any():
mask = unit_mask_image_fg
elif unit_mask_image is not None and (unit_mask_image > 5).any():
mask = unit_mask_image
elif unit_image_fg is not None and (unit_image_fg > 5).any():
mask = unit_image_fg
else:
mask = None
image = self.try_crop_image_with_a1111_mask(p, unit, image, resize_mode, preprocessor)
if mask is not None:
mask = cv2.resize(HWC3(mask), (image.shape[1], image.shape[0]), interpolation=cv2.INTER_NEAREST)
mask = self.try_crop_image_with_a1111_mask(p, unit, mask, resize_mode, preprocessor)
image_list = [[image, mask]]
if resize_mode == external_code.ResizeMode.OUTER_FIT and preprocessor.expand_mask_when_resize_and_fill:
new_image_list = []
for input_image, input_mask in image_list:
if input_mask is None:
input_mask = np.zeros_like(input_image)
input_mask = crop_and_resize_image(
input_mask,
external_code.ResizeMode.OUTER_FIT, h, w,
fill_border_with_255=True,
)
input_image = crop_and_resize_image(
input_image,
external_code.ResizeMode.OUTER_FIT, h, w,
fill_border_with_255=False,
)
new_image_list.append((input_image, input_mask))
image_list = new_image_list
return image_list, resize_mode
@staticmethod
def get_target_dimensions(p: StableDiffusionProcessing) -> Tuple[int, int, int, int]:
"""Returns (h, w, hr_h, hr_w)."""
h = align_dim_latent(p.height)
w = align_dim_latent(p.width)
high_res_fix = (
isinstance(p, StableDiffusionProcessingTxt2Img)
and getattr(p, 'enable_hr', False)
)
if high_res_fix:
if p.hr_resize_x == 0 and p.hr_resize_y == 0:
hr_y = int(p.height * p.hr_scale)
hr_x = int(p.width * p.hr_scale)
else:
hr_y, hr_x = p.hr_resize_y, p.hr_resize_x
hr_y = align_dim_latent(hr_y)
hr_x = align_dim_latent(hr_x)
else:
hr_y = h
hr_x = w
return h, w, hr_y, hr_x
@torch.no_grad()
def process_unit_after_click_generate(self,
p: StableDiffusionProcessing,
unit: ControlNetUnit,
params: ControlNetCachedParameters,
*args, **kwargs):
h, w, hr_y, hr_x = self.get_target_dimensions(p)
has_high_res_fix = (
isinstance(p, StableDiffusionProcessingTxt2Img)
and getattr(p, 'enable_hr', False)
)
if unit.use_preview_as_input:
unit.module = 'None'
preprocessor = global_state.get_preprocessor(unit.module)
input_list, resize_mode = self.get_input_data(p, unit, preprocessor, h, w)
preprocessor_outputs = []
control_masks = []
preprocessor_output_is_image = False
preprocessor_output = None
def optional_tqdm(iterable, use_tqdm):
from tqdm import tqdm
return tqdm(iterable) if use_tqdm else iterable
for input_image, input_mask in optional_tqdm(input_list, len(input_list) > 1):
if unit.pixel_perfect:
unit.processor_res = external_code.pixel_perfect_resolution(
input_image,
target_H=h,
target_W=w,
resize_mode=resize_mode,
)
seed = set_numpy_seed(p)
logger.debug(f"Use numpy seed {seed}.")
logger.info(f"Using preprocessor: {unit.module}")
logger.info(f'preprocessor resolution = {unit.processor_res}')
preprocessor_output = preprocessor(
input_image=input_image,
input_mask=input_mask,
resolution=unit.processor_res,
slider_1=unit.threshold_a,
slider_2=unit.threshold_b,
)
preprocessor_outputs.append(preprocessor_output)
preprocessor_output_is_image = judge_image_type(preprocessor_output)
if input_mask is not None:
control_masks.append(input_mask)
if len(input_list) > 1 and not preprocessor_output_is_image:
logger.info('Batch wise input only support controlnet, control-lora, and t2i adapters!')
break
if has_high_res_fix:
hr_option = HiResFixOption.from_value(unit.hr_option)
else:
hr_option = HiResFixOption.BOTH
alignment_indices = [i % len(preprocessor_outputs) for i in range(p.batch_size)]
def attach_extra_result_image(img: np.ndarray, is_high_res: bool = False):
if (
(is_high_res and hr_option.high_res_enabled) or
(not is_high_res and hr_option.low_res_enabled)
) and unit.save_detected_map:
p.extra_result_images.append(img)
if preprocessor_output_is_image:
params.control_cond = []
params.control_cond_for_hr_fix = []
for preprocessor_output in preprocessor_outputs:
control_cond = crop_and_resize_image(preprocessor_output, resize_mode, h, w)
attach_extra_result_image(external_code.visualize_inpaint_mask(control_cond))
params.control_cond.append(numpy_to_pytorch(control_cond).movedim(-1, 1))
params.control_cond = torch.cat(params.control_cond, dim=0)[alignment_indices].contiguous()
if has_high_res_fix:
for preprocessor_output in preprocessor_outputs:
control_cond_for_hr_fix = crop_and_resize_image(preprocessor_output, resize_mode, hr_y, hr_x)
attach_extra_result_image(external_code.visualize_inpaint_mask(control_cond_for_hr_fix), is_high_res=True)
params.control_cond_for_hr_fix.append(numpy_to_pytorch(control_cond_for_hr_fix).movedim(-1, 1))
params.control_cond_for_hr_fix = torch.cat(params.control_cond_for_hr_fix, dim=0)[alignment_indices].contiguous()
else:
params.control_cond_for_hr_fix = params.control_cond
else:
params.control_cond = preprocessor_output
params.control_cond_for_hr_fix = preprocessor_output
attach_extra_result_image(input_image)
if len(control_masks) > 0:
params.control_mask = []
params.control_mask_for_hr_fix = []
for input_mask in control_masks:
fill_border = preprocessor.fill_mask_with_one_when_resize_and_fill
control_mask = crop_and_resize_image(input_mask, resize_mode, h, w, fill_border)
attach_extra_result_image(control_mask)
control_mask = numpy_to_pytorch(control_mask).movedim(-1, 1)[:, :1]
params.control_mask.append(control_mask)
if has_high_res_fix:
control_mask_for_hr_fix = crop_and_resize_image(input_mask, resize_mode, hr_y, hr_x, fill_border)
attach_extra_result_image(control_mask_for_hr_fix, is_high_res=True)
control_mask_for_hr_fix = numpy_to_pytorch(control_mask_for_hr_fix).movedim(-1, 1)[:, :1]
params.control_mask_for_hr_fix.append(control_mask_for_hr_fix)
params.control_mask = torch.cat(params.control_mask, dim=0)[alignment_indices].contiguous()
if has_high_res_fix:
params.control_mask_for_hr_fix = torch.cat(params.control_mask_for_hr_fix, dim=0)[alignment_indices].contiguous()
else:
params.control_mask_for_hr_fix = params.control_mask
if preprocessor.do_not_need_model:
model_filename = 'Not Needed'
params.model = ControlModelPatcher()
else:
assert unit.model != 'None', 'You have not selected any control model!'
model_filename = global_state.get_controlnet_filename(unit.model)
params.model = cached_controlnet_loader(model_filename)
assert params.model is not None, logger.error(f"Recognizing Control Model failed: {model_filename}")
params.preprocessor = preprocessor
params.preprocessor.process_after_running_preprocessors(process=p, params=params, **kwargs)
params.model.process_after_running_preprocessors(process=p, params=params, **kwargs)
logger.info(f"Current ControlNet {type(params.model).__name__}: {model_filename}")
return
@torch.no_grad()
def process_unit_before_every_sampling(self,
p: StableDiffusionProcessing,
unit: ControlNetUnit,
params: ControlNetCachedParameters,
*args, **kwargs):
is_hr_pass = getattr(p, 'is_hr_pass', False)
has_high_res_fix = (
isinstance(p, StableDiffusionProcessingTxt2Img)
and getattr(p, 'enable_hr', False)
)
if has_high_res_fix:
hr_option = HiResFixOption.from_value(unit.hr_option)
else:
hr_option = HiResFixOption.BOTH
if has_high_res_fix and is_hr_pass and (not hr_option.high_res_enabled):
logger.info(f"ControlNet Skipped High-res pass.")
return
if has_high_res_fix and (not is_hr_pass) and (not hr_option.low_res_enabled):
logger.info(f"ControlNet Skipped Low-res pass.")
return
if is_hr_pass:
cond = params.control_cond_for_hr_fix
mask = params.control_mask_for_hr_fix
else:
cond = params.control_cond
mask = params.control_mask
kwargs.update(dict(
unit=unit,
params=params,
cond_original=cond.clone() if isinstance(cond, torch.Tensor) else cond,
mask_original=mask.clone() if isinstance(mask, torch.Tensor) else mask,
))
params.model.strength = float(unit.weight)
params.model.start_percent = float(unit.guidance_start)
params.model.end_percent = float(unit.guidance_end)
params.model.positive_advanced_weighting = None
params.model.negative_advanced_weighting = None
params.model.advanced_frame_weighting = None
params.model.advanced_sigma_weighting = None
soft_weighting = {
'input': [0.09941396206337118, 0.12050177219802567, 0.14606275417942507, 0.17704576264172736,
0.214600924414215,
0.26012233262329093, 0.3152997971191405, 0.3821815722656249, 0.4632503906249999, 0.561515625,
0.6806249999999999, 0.825],
'middle': [0.561515625] if p.sd_model.is_sdxl else [1.0],
'output': [0.09941396206337118, 0.12050177219802567, 0.14606275417942507, 0.17704576264172736,
0.214600924414215,
0.26012233262329093, 0.3152997971191405, 0.3821815722656249, 0.4632503906249999, 0.561515625,
0.6806249999999999, 0.825]
}
zero_weighting = {
'input': [0.0] * 12,
'middle': [0.0],
'output': [0.0] * 12
}
if unit.control_mode == external_code.ControlMode.CONTROL.value:
params.model.positive_advanced_weighting = soft_weighting.copy()
params.model.negative_advanced_weighting = zero_weighting.copy()
if unit.control_mode == external_code.ControlMode.PROMPT.value:
params.model.positive_advanced_weighting = soft_weighting.copy()
params.model.negative_advanced_weighting = soft_weighting.copy()
if is_hr_pass and params.preprocessor.use_soft_projection_in_hr_fix:
params.model.positive_advanced_weighting = soft_weighting.copy()
params.model.negative_advanced_weighting = soft_weighting.copy()
cond, mask = params.preprocessor.process_before_every_sampling(p, cond, mask, *args, **kwargs)
params.model.advanced_mask_weighting = mask
params.model.process_before_every_sampling(p, cond, mask, *args, **kwargs)
logger.info(f"ControlNet Method {params.preprocessor.name} patched.")
return
@staticmethod
def bound_check_params(unit: ControlNetUnit) -> None:
"""
Checks and corrects negative parameters in ControlNetUnit 'unit'.
Parameters 'processor_res', 'threshold_a', 'threshold_b' are reset to
their default values if negative.
Args:
unit (ControlNetUnit): The ControlNetUnit instance to check.
"""
preprocessor = global_state.get_preprocessor(unit.module)
if unit.processor_res < 0:
unit.processor_res = int(preprocessor.slider_resolution.gradio_update_kwargs.get('value', 512))
if unit.threshold_a < 0:
unit.threshold_a = int(preprocessor.slider_1.gradio_update_kwargs.get('value', 1.0))
if unit.threshold_b < 0:
unit.threshold_b = int(preprocessor.slider_2.gradio_update_kwargs.get('value', 1.0))
return
@torch.no_grad()
def process_unit_after_every_sampling(self,
p: StableDiffusionProcessing,
unit: ControlNetUnit,
params: ControlNetCachedParameters,
*args, **kwargs):
params.preprocessor.process_after_every_sampling(p, params, *args, **kwargs)
params.model.process_after_every_sampling(p, params, *args, **kwargs)
return
@torch.no_grad()
def process(self, p, *args, **kwargs):
self.current_params = {}
enabled_units = self.get_enabled_units(args)
Infotext.write_infotext(enabled_units, p)
for i, unit in enumerate(enabled_units):
self.bound_check_params(unit)
params = ControlNetCachedParameters()
self.process_unit_after_click_generate(p, unit, params, *args, **kwargs)
self.current_params[i] = params
return
@torch.no_grad()
def process_before_every_sampling(self, p, *args, **kwargs):
for i, unit in enumerate(self.get_enabled_units(args)):
self.process_unit_before_every_sampling(p, unit, self.current_params[i], *args, **kwargs)
return
@torch.no_grad()
def postprocess_batch_list(self, p, pp, *args, **kwargs):
for i, unit in enumerate(self.get_enabled_units(args)):
self.process_unit_after_every_sampling(p, unit, self.current_params[i], pp, *args, **kwargs)
return
def postprocess(self, p, processed, *args):
self.current_params = {}
return
def on_ui_settings():
section = ('control_net', "ControlNet")
shared.opts.add_option("control_net_detectedmap_dir", shared.OptionInfo(
"detected_maps", "Directory for detected maps auto saving", section=section))
shared.opts.add_option("control_net_models_path", shared.OptionInfo(
"", "Extra path to scan for ControlNet models (e.g. training output directory)", section=section))
shared.opts.add_option("control_net_modules_path", shared.OptionInfo(
"",
"Path to directory containing annotator model directories (requires restart, overrides corresponding command line flag)",
section=section))
shared.opts.add_option("control_net_unit_count", shared.OptionInfo(
3, "Multi-ControlNet: ControlNet unit number (requires restart)", gr.Slider,
{"minimum": 1, "maximum": 10, "step": 1}, section=section))
shared.opts.add_option("control_net_model_cache_size", shared.OptionInfo(
5, "Model cache size (requires restart)", gr.Slider, {"minimum": 1, "maximum": 10, "step": 1}, section=section))
shared.opts.add_option("control_net_no_detectmap", shared.OptionInfo(
False, "Do not append detectmap to output", gr.Checkbox, {"interactive": True}, section=section))
shared.opts.add_option("control_net_detectmap_autosaving", shared.OptionInfo(
False, "Allow detectmap auto saving", gr.Checkbox, {"interactive": True}, section=section))
shared.opts.add_option("control_net_allow_script_control", shared.OptionInfo(
False, "Allow other script to control this extension", gr.Checkbox, {"interactive": True}, section=section))
shared.opts.add_option("control_net_sync_field_args", shared.OptionInfo(
True, "Paste ControlNet parameters in infotext", gr.Checkbox, {"interactive": True}, section=section))
shared.opts.add_option("controlnet_show_batch_images_in_ui", shared.OptionInfo(
False, "Show batch images in gradio gallery output", gr.Checkbox, {"interactive": True}, section=section))
shared.opts.add_option("controlnet_increment_seed_during_batch", shared.OptionInfo(
False, "Increment seed after each controlnet batch iteration", gr.Checkbox, {"interactive": True},
section=section))
shared.opts.add_option("controlnet_disable_openpose_edit", shared.OptionInfo(
False, "Disable openpose edit", gr.Checkbox, {"interactive": True}, section=section))
shared.opts.add_option("controlnet_disable_photopea_edit", shared.OptionInfo(
False, "Disable photopea edit", gr.Checkbox, {"interactive": True}, section=section))
shared.opts.add_option("controlnet_photopea_warning", shared.OptionInfo(
True, "Photopea popup warning", gr.Checkbox, {"interactive": True}, section=section))
shared.opts.add_option("controlnet_input_thumbnail", shared.OptionInfo(
True, "Input image thumbnail on unit header", gr.Checkbox, {"interactive": True}, section=section))
script_callbacks.on_ui_settings(on_ui_settings)
script_callbacks.on_infotext_pasted(Infotext.on_infotext_pasted)
script_callbacks.on_after_component(ControlNetUiGroup.on_after_component)
script_callbacks.on_before_reload(ControlNetUiGroup.reset)
script_callbacks.on_app_started(controlnet_api)

449
extensions-builtin/sd_forge_controlnet/scripts/xyz_grid_support.py Executable file
View File

@@ -0,0 +1,449 @@
import re
import numpy as np
from modules import scripts, shared
try:
from lib_controlnet.global_state import update_controlnet_filenames, cn_models_names, get_preprocessor_names
from lib_controlnet.external_code import ResizeMode, ControlMode
except (ImportError, NameError):
import_error = True
else:
import_error = False
DEBUG_MODE = False
def debug_info(func):
def debug_info_(*args, **kwargs):
if DEBUG_MODE:
print(f"Debug info: {func.__name__}, {args}")
return func(*args, **kwargs)
return debug_info_
def find_dict(dict_list, keyword, search_key="name", stop=False):
result = next((d for d in dict_list if d[search_key] == keyword), None)
if result or not stop:
return result
else:
raise ValueError(f"Dictionary with value '{keyword}' in key '{search_key}' not found.")
def flatten(lst):
result = []
for element in lst:
if isinstance(element, list):
result.extend(flatten(element))
else:
result.append(element)
return result
def is_all_included(target_list, check_list, allow_blank=False, stop=False):
for element in flatten(target_list):
if allow_blank and str(element) in ["None", ""]:
continue
elif element not in check_list:
if not stop:
return False
else:
raise ValueError(f"'{element}' is not included in check list.")
return True
class ListParser():
"""This class restores a broken list caused by the following process
in the xyz_grid module.
-> valslist = [x.strip() for x in chain.from_iterable(
csv.reader(StringIO(vals)))]
It also performs type conversion,
adjusts the number of elements in the list, and other operations.
This class directly modifies the received list.
"""
numeric_pattern = {
int: {
"range": r"\s*([+-]?\s*\d+)\s*-\s*([+-]?\s*\d+)(?:\s*\(([+-]\d+)\s*\))?\s*",
"count": r"\s*([+-]?\s*\d+)\s*-\s*([+-]?\s*\d+)(?:\s*\[(\d+)\s*\])?\s*"
},
float: {
"range": r"\s*([+-]?\s*\d+(?:\.\d*)?)\s*-\s*([+-]?\s*\d+(?:\.\d*)?)(?:\s*\(([+-]\d+(?:\.\d*)?)\s*\))?\s*",
"count": r"\s*([+-]?\s*\d+(?:\.\d*)?)\s*-\s*([+-]?\s*\d+(?:\.\d*)?)(?:\s*\[(\d+(?:\.\d*)?)\s*\])?\s*"
}
}
################################################
#
# Initialization method from here.
#
################################################
def __init__(self, my_list, converter=None, allow_blank=True, exclude_list=None, run=True):
self.my_list = my_list
self.converter = converter
self.allow_blank = allow_blank
self.exclude_list = exclude_list
self.re_bracket_start = None
self.re_bracket_start_precheck = None
self.re_bracket_end = None
self.re_bracket_end_precheck = None
self.re_range = None
self.re_count = None
self.compile_regex()
if run:
self.auto_normalize()
def compile_regex(self):
exclude_pattern = "|".join(self.exclude_list) if self.exclude_list else None
if exclude_pattern is None:
self.re_bracket_start = re.compile(r"^\[")
self.re_bracket_end = re.compile(r"\]$")
else:
self.re_bracket_start = re.compile(fr"^\[(?!(?:{exclude_pattern})\])")
self.re_bracket_end = re.compile(fr"(?<!\[(?:{exclude_pattern}))\]$")
if self.converter not in self.numeric_pattern:
return self
# If the converter is either int or float.
self.re_range = re.compile(self.numeric_pattern[self.converter]["range"])
self.re_count = re.compile(self.numeric_pattern[self.converter]["count"])
self.re_bracket_start_precheck = None
self.re_bracket_end_precheck = self.re_count
return self
################################################
#
# Public method from here.
#
################################################
################################################
# This method is executed at the time of initialization.
#
def auto_normalize(self):
if not self.has_list_notation():
self.numeric_range_parser()
self.type_convert()
return self
else:
self.fix_structure()
self.numeric_range_parser()
self.type_convert()
self.fill_to_longest()
return self
def has_list_notation(self):
return any(self._search_bracket(s) for s in self.my_list)
def numeric_range_parser(self, my_list=None, depth=0):
if self.converter not in self.numeric_pattern:
return self
my_list = self.my_list if my_list is None else my_list
result = []
is_matched = False
for s in my_list:
if isinstance(s, list):
result.extend(self.numeric_range_parser(s, depth+1))
continue
match = self._numeric_range_to_list(s)
if s != match:
is_matched = True
result.extend(match if not depth else [match])
continue
else:
result.append(s)
continue
if depth:
return self._transpose(result) if is_matched else [result]
else:
my_list[:] = result
return self
def type_convert(self, my_list=None):
my_list = self.my_list if my_list is None else my_list
for i, s in enumerate(my_list):
if isinstance(s, list):
self.type_convert(s)
elif self.allow_blank and (str(s) in ["None", ""]):
my_list[i] = None
elif self.converter:
my_list[i] = self.converter(s)
else:
my_list[i] = s
return self
def fix_structure(self):
def is_same_length(list1, list2):
return len(list1) == len(list2)
start_indices, end_indices = [], []
for i, s in enumerate(self.my_list):
if is_same_length(start_indices, end_indices):
replace_string = self._search_bracket(s, "[", replace="")
if s != replace_string:
s = replace_string
start_indices.append(i)
if not is_same_length(start_indices, end_indices):
replace_string = self._search_bracket(s, "]", replace="")
if s != replace_string:
s = replace_string
end_indices.append(i + 1)
self.my_list[i] = s
if not is_same_length(start_indices, end_indices):
raise ValueError(f"Lengths of {start_indices} and {end_indices} are different.")
# Restore the structure of a list.
for i, j in zip(reversed(start_indices), reversed(end_indices)):
self.my_list[i:j] = [self.my_list[i:j]]
return self
def fill_to_longest(self, my_list=None, value=None, index=None):
my_list = self.my_list if my_list is None else my_list
if not self.sublist_exists(my_list):
return self
max_length = max(len(sub_list) for sub_list in my_list if isinstance(sub_list, list))
for i, sub_list in enumerate(my_list):
if isinstance(sub_list, list):
fill_value = value if index is None else sub_list[index]
my_list[i] = sub_list + [fill_value] * (max_length-len(sub_list))
return self
def sublist_exists(self, my_list=None):
my_list = self.my_list if my_list is None else my_list
return any(isinstance(item, list) for item in my_list)
def all_sublists(self, my_list=None): # Unused method
my_list = self.my_list if my_list is None else my_list
return all(isinstance(item, list) for item in my_list)
def get_list(self): # Unused method
return self.my_list
################################################
#
# Private method from here.
#
################################################
def _search_bracket(self, string, bracket="[", replace=None):
if bracket == "[":
pattern = self.re_bracket_start
precheck = self.re_bracket_start_precheck # None
elif bracket == "]":
pattern = self.re_bracket_end
precheck = self.re_bracket_end_precheck
else:
raise ValueError(f"Invalid argument provided. (bracket: {bracket})")
if precheck and precheck.fullmatch(string):
return None if replace is None else string
elif replace is None:
return pattern.search(string)
else:
return pattern.sub(replace, string)
def _numeric_range_to_list(self, string):
match = self.re_range.fullmatch(string)
if match is not None:
if self.converter == int:
start = int(match.group(1))
end = int(match.group(2)) + 1
step = int(match.group(3)) if match.group(3) is not None else 1
return list(range(start, end, step))
else: # float
start = float(match.group(1))
end = float(match.group(2))
step = float(match.group(3)) if match.group(3) is not None else 1
return np.arange(start, end + step, step).tolist()
match = self.re_count.fullmatch(string)
if match is not None:
if self.converter == int:
start = int(match.group(1))
end = int(match.group(2))
num = int(match.group(3)) if match.group(3) is not None else 1
return [int(x) for x in np.linspace(start=start, stop=end, num=num).tolist()]
else: # float
start = float(match.group(1))
end = float(match.group(2))
num = int(match.group(3)) if match.group(3) is not None else 1
return np.linspace(start=start, stop=end, num=num).tolist()
return string
def _transpose(self, my_list=None):
my_list = self.my_list if my_list is None else my_list
my_list = [item if isinstance(item, list) else [item] for item in my_list]
self.fill_to_longest(my_list, index=-1)
return np.array(my_list, dtype=object).T.tolist()
################################################
#
# The methods of ListParser class end here.
#
################################################
################################################################
################################################################
#
# Starting the main process of this module.
#
# functions are executed in this order:
# find_module
# add_axis_options
# identity
# enable_script_control
# apply_field
# confirm
# bool_
# choices_for
# make_excluded_list
# config lists for AxisOptions:
# validation_data
# extra_axis_options
################################################################
################################################################
def find_module(module_names):
if isinstance(module_names, str):
module_names = [s.strip() for s in module_names.split(",")]
for data in scripts.scripts_data:
if data.script_class.__module__ in module_names and hasattr(data, "module"):
return data.module
return None
def add_axis_options(xyz_grid):
################################################
#
# Define a function to pass to the AxisOption class from here.
#
################################################
################################################
# Set this function as the type attribute of the AxisOption class.
# To skip the following processing of xyz_grid module.
# -> valslist = [opt.type(x) for x in valslist]
# Perform type conversion using the function
# set to the confirm attribute instead.
#
def identity(x):
return x
def enable_script_control():
shared.opts.data["control_net_allow_script_control"] = True
def apply_field(field):
@debug_info
def apply_field_(p, x, xs):
enable_script_control()
setattr(p, field, x)
return apply_field_
################################################
# The confirm function defined in this module
# enables list notation and performs type conversion.
#
# Example:
# any = [any, any, any, ...]
# [any] = [any, None, None, ...]
# [None, None, any] = [None, None, any]
# [,,any] = [None, None, any]
# any, [,any,] = [any, any, any, ...], [None, any, None]
#
# Enabled Only:
# any = [any] = [any, None, None, ...]
# (any and [any] are considered equivalent)
#
def confirm(func_or_str):
@debug_info
def confirm_(p, xs):
if callable(func_or_str): # func_or_str is converter
ListParser(xs, func_or_str, allow_blank=True)
return
elif isinstance(func_or_str, str): # func_or_str is keyword
valid_data = find_dict(validation_data, func_or_str, stop=True)
converter = valid_data["type"]
exclude_list = valid_data["exclude"]() if valid_data["exclude"] else None
check_list = valid_data["check"]()
ListParser(xs, converter, allow_blank=True, exclude_list=exclude_list)
is_all_included(xs, check_list, allow_blank=True, stop=True)
return
else:
raise TypeError(f"Argument must be callable or str, not {type(func_or_str).__name__}.")
return confirm_
def bool_(string):
string = str(string)
if string in ["None", ""]:
return None
elif string.lower() in ["true", "1"]:
return True
elif string.lower() in ["false", "0"]:
return False
else:
raise ValueError(f"Could not convert string to boolean: {string}")
def choices_bool():
return ["False", "True"]
def choices_model():
update_controlnet_filenames()
return list(cn_models_names.values())
def choices_control_mode():
return [e.value for e in ControlMode]
def choices_resize_mode():
return [e.value for e in ResizeMode]
def choices_preprocessor():
return list(get_preprocessor_names())
def make_excluded_list():
pattern = re.compile(r"\[(\w+)\]")
return [match.group(1) for s in choices_model()
for match in pattern.finditer(s)]
validation_data = [
{"name": "model", "type": str, "check": choices_model, "exclude": make_excluded_list},
{"name": "control_mode", "type": str, "check": choices_control_mode, "exclude": None},
{"name": "resize_mode", "type": str, "check": choices_resize_mode, "exclude": None},
{"name": "preprocessor", "type": str, "check": choices_preprocessor, "exclude": None},
]
extra_axis_options = [
xyz_grid.AxisOption("[ControlNet] Enabled", identity, apply_field("control_net_enabled"), confirm=confirm(bool_), choices=choices_bool),
xyz_grid.AxisOption("[ControlNet] Model", identity, apply_field("control_net_model"), confirm=confirm("model"), choices=choices_model, cost=0.9),
xyz_grid.AxisOption("[ControlNet] Weight", identity, apply_field("control_net_weight"), confirm=confirm(float)),
xyz_grid.AxisOption("[ControlNet] Guidance Start", identity, apply_field("control_net_guidance_start"), confirm=confirm(float)),
xyz_grid.AxisOption("[ControlNet] Guidance End", identity, apply_field("control_net_guidance_end"), confirm=confirm(float)),
xyz_grid.AxisOption("[ControlNet] Control Mode", identity, apply_field("control_net_control_mode"), confirm=confirm("control_mode"), choices=choices_control_mode),
xyz_grid.AxisOption("[ControlNet] Resize Mode", identity, apply_field("control_net_resize_mode"), confirm=confirm("resize_mode"), choices=choices_resize_mode),
xyz_grid.AxisOption("[ControlNet] Preprocessor", identity, apply_field("control_net_module"), confirm=confirm("preprocessor"), choices=choices_preprocessor),
xyz_grid.AxisOption("[ControlNet] Pre Resolution", identity, apply_field("control_net_pres"), confirm=confirm(int)),
xyz_grid.AxisOption("[ControlNet] Pre Threshold A", identity, apply_field("control_net_pthr_a"), confirm=confirm(float)),
xyz_grid.AxisOption("[ControlNet] Pre Threshold B", identity, apply_field("control_net_pthr_b"), confirm=confirm(float)),
]
xyz_grid.axis_options.extend(extra_axis_options)
def run():
xyz_grid = find_module("xyz_grid.py, xy_grid.py")
if xyz_grid:
add_axis_options(xyz_grid)
if not import_error:
run()

251
extensions-builtin/sd_forge_controlnet/style.css Executable file
View File

@@ -0,0 +1,251 @@
/* InputAccordion alignment */
/* Flex container */
.controlnet .svelte-vt1mxs {
display: flex;
flex-wrap: wrap;
flex-direction: row;
gap: 10px;
/* Adjusts the space between items */
}
.controlnet .input-accordion {
flex: 1 1 calc(50% - 10px);
/* Adjusts for the gap, default 2 columns */
display: flex;
align-items: center;
}
/* Media query for screens smaller than a specific width */
@media (max-width: 600px) {
.controlnet .input-accordion {
flex: 1 1 100%;
/* Changes to 1 column when window width is ≤ 600px */
}
}
/* Input image thumbnail */
.cnet-thumbnail {
height: 3rem !important;
border: 1px solid var(--button-secondary-border-color);
}
.controlnet .input-accordion .label-wrap>span:nth-child(1) {
display: flex;
flex-direction: row;
align-items: center;
gap: 5px;
}
.controlnet .input-accordion .icon {
height: 1rem;
width: 1rem;
}
.controlnet .input-accordion .label-wrap {
align-items: center;
}
.cnet-modal {
display: none;
/* Hidden by default */
position: fixed;
/* Stay in place */
z-index: 2147483647;
/* Sit on top */
left: 0;
top: 0;
width: 100%;
/* Full width */
height: 100%;
/* Full height */
overflow: auto;
/* Enable scroll if needed */
background-color: rgba(0, 0, 0, 0.4);
/* Black with opacity */
max-width: none !important;
/* Fix sizing with SD.Next (vladmandic/automatic#2594) */
}
.cnet-modal-content {
position: relative;
background-color: var(--background-fill-primary);
margin: 5vh auto;
/* 15% from the top and centered */
padding: 20px;
border: 1px solid #888;
width: 95%;
height: 90vh;
/* Could be more or less, depending on screen size */
box-shadow: 0 4px 8px 0 rgba(0, 0, 0, 0.2), 0 6px 20px 0 rgba(0, 0, 0, 0.19);
animation-name: animatetop;
animation-duration: 0.4s;
max-width: none !important;
/* Fix sizing with SD.Next (vladmandic/automatic#2594) */
}
.cnet-modal-content iframe {
position: absolute;
top: 0;
left: 0;
width: 100%;
height: 100%;
border: none;
}
.cnet-modal-content.alert {
padding: var(--size-5);
}
.cnet-modal-content.alert ul {
list-style-type: none;
}
.cnet-modal-close {
color: white !important;
right: 0.25em;
top: 0;
cursor: pointer;
position: absolute;
font-size: 56px;
font-weight: bold;
}
@keyframes animatetop {
from {
top: -300px;
opacity: 0
}
to {
top: 0;
opacity: 1
}
}
.cnet-generated-image-control-group,
.cnet-upload-pose {
display: flex;
flex-direction: column;
align-items: flex-end;
position: absolute;
right: var(--size-2);
bottom: var(--size-2);
}
/* Gradio button style */
.cnet-download-pose a,
.cnet-close-preview,
.cnet-edit-pose,
.cnet-upload-pose,
.cnet-photopea-child-trigger {
font-size: x-small !important;
font-weight: bold !important;
padding: 2px !important;
box-shadow: var(--shadow-drop);
border: 1px solid var(--button-secondary-border-color);
border-radius: var(--radius-sm);
background: var(--background-fill-primary);
height: var(--size-5);
color: var(--block-label-text-color) !important;
display: flex;
justify-content: center;
cursor: pointer;
}
.cnet-download-pose:hover a,
.cnet-close-preview:hover a,
.cnet-edit-pose:hover,
.cnet-upload-pose:hover,
.cnet-photopea-child-trigger:hover {
color: var(--block-label-text-color) !important;
}
.cnet-unit-active {
color: green !important;
font-weight: bold !important;
}
.dark .cnet-unit-active {
color: greenyellow !important;
}
.cnet-badge {
display: inline-block;
padding: 0.25em 0.75em;
font-size: 0.75em;
font-weight: bold;
color: white;
border-radius: 0.5em;
text-align: center;
vertical-align: middle;
margin-left: var(--size-2);
}
.cnet-badge.primary {
background-color: green;
}
.cnet-a1111-badge {
position: absolute;
bottom: 0px;
right: 0px;
}
.cnet-disabled-radio {
opacity: 50%;
}
.controlnet_row {
margin-top: 10px !important;
}
/* JSON pose upload button styling */
.cnet-upload-pose input[type=file] {
position: absolute;
left: 0;
top: 0;
opacity: 0;
width: 100%;
height: 100%;
}
/* Photopea integration styles */
.photopea-button-group {
position: absolute;
top: -30px; /* 20px modal padding + 10px margin */
}
.photopea-button {
font-size: 3rem;
font-weight: bold;
padding: 2px !important;
margin: 2px !important;
box-shadow: var(--shadow-drop);
border: 1px solid var(--button-secondary-border-color);
border-radius: var(--radius-sm);
background: var(--background-fill-primary);
color: var(--block-label-text-color);
}
.controlnet_control_type_filter_group label {
background: unset !important;
border: unset !important;
margin-left: -10px !important;
}
.controlnet_control_type_filter_group > span {
display: none !important;
}
.controlnet_control_type_filter_group > .wrap {
margin-top: -20px !important;
}
.cnet-toolbutton {
background: unset !important;
border: unset !important;
}
.range-slider {
margin-top: -8px;
}