Update ImHex to latest, preserving additions#1
Open
the-claw wants to merge 423 commits into
Open
Conversation
for the ghost break point bug
<!-- Please provide as much information as possible about what your PR aims to do. PRs with no description will most likely be closed until more information is provided. If you're planing on changing fundamental behaviour or add big new features, please open a GitHub Issue first before starting to work on it. If it's not something big and you still want to contact us about it, feel free to do so ! --> ### Problem description <!-- Describe the bug that you fixed/feature request that you implemented, or link to an existing issue describing it --> ### Implementation description <!-- Explain what you did to correct the problem --> ### Screenshots <!-- If your change is visual, take a screenshot showing it. Ideally, make before/after sceenshots --> ### Additional things <!-- Anything else you would like to say -->
…olv#2574) <!-- Please provide as much information as possible about what your PR aims to do. PRs with no description will most likely be closed until more information is provided. If you're planing on changing fundamental behaviour or add big new features, please open a GitHub Issue first before starting to work on it. If it's not something big and you still want to contact us about it, feel free to do so ! --> ### Problem description <!-- Describe the bug that you fixed/feature request that you implemented, or link to an existing issue describing it --> ### Implementation description <!-- Explain what you did to correct the problem --> ### Screenshots <!-- If your change is visual, take a screenshot showing it. Ideally, make before/after sceenshots --> ### Additional things <!-- Anything else you would like to say -->
(cherry picked from commit 5a85356)
<!-- Please provide as much information as possible about what your PR aims to do. PRs with no description will most likely be closed until more information is provided. If you're planing on changing fundamental behaviour or add big new features, please open a GitHub Issue first before starting to work on it. If it's not something big and you still want to contact us about it, feel free to do so ! --> ### Problem description <!-- Describe the bug that you fixed/feature request that you implemented, or link to an existing issue describing it --> ### Implementation description <!-- Explain what you did to correct the problem --> ### Screenshots <!-- If your change is visual, take a screenshot showing it. Ideally, make before/after sceenshots --> ### Additional things <!-- Anything else you would like to say -->
(cherry picked from commit b78a102)
(cherry picked from commit 8e4ccef)
(cherry picked from commit 76cf877)
(cherry picked from commit 34bc55a)
<!-- Please provide as much information as possible about what your PR aims to do. PRs with no description will most likely be closed until more information is provided. If you're planing on changing fundamental behaviour or add big new features, please open a GitHub Issue first before starting to work on it. If it's not something big and you still want to contact us about it, feel free to do so ! --> ### Problem description <!-- Describe the bug that you fixed/feature request that you implemented, or link to an existing issue describing it --> ### Implementation description <!-- Explain what you did to correct the problem --> ### Screenshots <!-- If your change is visual, take a screenshot showing it. Ideally, make before/after sceenshots --> ### Additional things <!-- Anything else you would like to say -->
The problem is caused by calculating space used by strings using string.length() which only works if every char in the string is encoded in 1 byte. The wrongly calculated space is then used to truncate the string at the wrong place. Finally, the wrongly calculated space is used to align the log entries which results in misalignment. The fix consists on counting utf-8 characters to calculate the spaces and using a function designed to work on utf-8 strings that returns sub-stings. A further problem with alignment was caused when using CJK chars that are wider than regular half-width characters which throws the space calculations off. This was fixed by detecting when CJK glyphs are being used and adjusting the sizes using a reasonable approximation that 1 CJK char ~ 1.75 half-width chars. Compare the two images. Before the PR: <img width="1665" height="858" alt="zoomit" src="https://github.com/user-attachments/assets/9d1a467d-a0fa-4cf6-882f-51712044234f" /> and after the PR: <img width="1925" height="863" alt="after" src="https://github.com/user-attachments/assets/bb905f79-8bb5-4bd3-85b5-fb36327bc19b" /> Note how the utf-8 strings are truncated more than they should and that for those same entries the alignment is lost but are not with the changes. The image shows how CJK glyphs are handled.
…erWolv#2718) <!-- Please provide as much information as possible about what your PR aims to do. PRs with no description will most likely be closed until more information is provided. If you're planing on changing fundamental behaviour or add big new features, please open a GitHub Issue first before starting to work on it. If it's not something big and you still want to contact us about it, feel free to do so ! --> ### Problem description Commit WerWolv@f813656 introduced a nullptr dereference that happens on the first frame of a provider being opened. ``` [12:03:47] [INFO] [main | Main] Switching to workspace 'Default' Process 89715 stopped * thread #1, name = 'ImHex 🔍', stop reason = signal SIGSEGV: address not mapped to object (fault address=0x0) frame #0: 0x00007fffcdd43958 ui.hexpluglib`::drawEditor() at hex_editor.cpp:648:46 645 ImGui::TableSetupScrollFreeze(0, 2); 646 647 // Row address column -> 648 u64 maxAddress = m_provider->getActualSize(); ^ 649 if (maxAddress > 0) 650 maxAddress--; 651 if ((m_scrollPosition + m_visibleRowCount) * bytesPerRow < maxAddress) (lldb) bt * thread #1, name = 'ImHex 🔍', stop reason = signal SIGSEGV: address not mapped to object (fault address=0x0) * frame #0: 0x00007fffcdd43958 ui.hexpluglib`::drawEditor() at hex_editor.cpp:648:46 frame #1: 0x00007fffcdd4954b ui.hexpluglib`::draw() at hex_editor.cpp:1409:15 frame WerWolv#2: 0x00007fffdc01eb21 diffing.hexplug`::drawDiffColumn() at view_diff.cpp:64:30 frame WerWolv#3: 0x00007fffdc01d56d diffing.hexplug`::drawContent() at view_diff.cpp:271:28 frame WerWolv#4: 0x00007ffff7af7413 libimhex.so.1.39.0.WIP`::draw() at view.cpp:201:23 frame WerWolv#5: 0x0000555555593e7d imhex`::frame() at window.cpp:759:23 frame WerWolv#6: 0x000055555559237d imhex`hex::Window::fullFrame() at window.cpp:200:19 frame WerWolv#7: 0x00005555555948b5 imhex`::loop() at window.cpp:297:19 frame WerWolv#8: 0x00005555555c211c imhex`::runImHex() at desktop.cpp:54:28 frame WerWolv#9: 0x00005555555836a6 imhex`main at main.cpp:73:12 frame WerWolv#10: 0x00007ffff682f43b libc.so.6`___lldb_unnamed_symbol_2b3c0 + 123 frame WerWolv#11: 0x00007ffff682f4eb libc.so.6`__libc_start_main + 139 frame WerWolv#12: 0x00005555555831d5 imhex`_start + 37 ``` ### Implementation description If nullptr check guard.
{fmt} library can deal with all the issues caused by utf-8 chars in
logs. Original code used {fmt} but only to pad the spaces calculated
assuming 1 byte per char.
Also included code to change the windows console to use utf-8 and named
the syntax highlighting task like other tasks created near it.
The other half of the issue was that the structures that the data inspector was using to store the 16 bit values were using a 32 bit underlying type instead of a 16 bit. That worked ok for little endian, but with 4 bytes the resulting value was actually the next u16 after the one selected.
The feature that selects and highlight's matching delimiters stores the line and column for the match. If you delete the matching delimiter while the highlighting is on, then the code was trying to access a non-existent line causing an exception. Fixed a bug that didn't assign the right colors when a custom types, a global placed variable and a member of that custom type shared the same name. Some names were changed. Some other code was optimized.
<!-- Please provide as much information as possible about what your PR aims to do. PRs with no description will most likely be closed until more information is provided. If you're planing on changing fundamental behaviour or add big new features, please open a GitHub Issue first before starting to work on it. If it's not something big and you still want to contact us about it, feel free to do so ! --> ### Problem description When launching ImHex for the first time, the welcome screen has a small typo where there's a missing space between "for" and "Reverse Engineers" <!-- Describe the bug that you fixed/feature request that you implemented, or link to an existing issue describing it --> ### Implementation description I looked up the string definition in the source code, found that it's located here: https://github.com/WerWolv/ImHex/blob/3a8f19cc90b60176b6d35fb26c7d50451555b6b3/plugins/builtin/source/content/out_of_box_experience.cpp#L112 and fixed the typo <!-- Explain what you did to correct the problem --> ### Screenshots <!-- If your change is visual, take a screenshot showing it. Ideally, make before/after sceenshots --> <img width="1274" height="716" alt="image" src="https://github.com/user-attachments/assets/30841caf-7905-41d0-b86f-c9fa47f9833e" /> Co-authored-by: paxcut <53811119+paxcut@users.noreply.github.com>
Doing
```cpp
{
}
```
or
```cpp
{
{
}
}
```
was making imhex crash when closing folds. The cases were not considered
during development because they are illegal pattern language
statements.fix was to adjust the code to accommodate for them.
…pups. (WerWolv#2738) Don't open the pattern suggestion popup if there are no suggestions or if a pattern is already loaded for that provider.
Both Devil May Cry HD Collection.tar and AC Unity.tar fail to download in the Content Store because of the spaces in the file names. Apparently curl will generate errors if the url link contains spaces. the fix is to url encode the spaces only, a process known as 'urlify'ing the link. The implementation uses boost regex replace to perform the substitution when the link is added to the entry. Perhaps a better fix would be to patch the python code that generates the Store response, but I don't have merge permissions on that repository and all the testing I have done indicates that the problem has been resolved.
Fixes issue WerWolv#2745. White spaces were missing when syntactic colors and show white spaces were both off. Also a crash could occur when adding new lines at the end of the file.
Changed the download parameters to improve the rate of failure but if it doesn't download we skip the no-gpu package generation and continue the build. I looked for the self-hosted version on werwolv.net but there is no downloads folder there anymore.
Allow skipping moving the windows No-GPU version without making the nightly build fail.
This attempts to fix the problems described in issue WerWolv#2748. When a selection is deleted, the selection ends are changed correctly, but the interactive selection is not so attempting to use it after the deletion produces unexpected results. The fix is to update the interactive selection as well. I also made some code more efficient by removing the creation of temporary objects that were not necessary. Some code was repeated several times, so I move the code into their own function. Cleaned the code handling mouse clicks and added some `using` typedefs to shorten code.
The exact same error occurring while building plutosvg has been fixed on the last VS 2026 which is 2 days old. Since we are still using 2022 we have to use the published workaround of disabling some optimization. We can't edit plutosvg code so the optimization needs to be disabled for the whole build.
Archlinux needs to change some headers to include `fmt/format.hpp` instead of `fmt/core.hpp`. The archlinux build will not be fixed until the PR in disassembler repo gets merged by somebody with push permissions there. MSVC builds had last change reverted because not only it did not fix the build errors, but it added some extra ones. Seeing how this was the only workaround we can apply we probably will need to update the compiler to 2026 and wait for the version that has the fix to be used in github actions. Weirdly enough, both msvc builds went through, so maybe the ICE only occurs sometimes.
fixes issue WerWolv#2758. Problem was caused by the improper parsing of the template arguments during the association of the variable type with the variable name. The fix uses the fact that template arguments containing comparison or shift operators must be enclosed with parentheses to avoid syntactic ambiguity, so when parentheses are detected they are used to skip parsing of what they surround. There was a lot of code duplication in the code that looked for matching delimiter which was removed and the detection and handling of the separate delimiter pairs was simplified greatly. During the time it took to create this PR other problems in the pattern editor were discovered and their fixes are also included. * The matching delimiter for templates was also set incorrectly when operators that could cause syntactic ambiguity were used. * The block identification line was drawn when cursor was near the block delimiters instead of anywhere within the block. The code blocks are slightly different from code folds (e.g. imports and comments) and correspond to indentation level so m_globalBlocks was renamed to m_indentBlocks and its generation separated from the generation of code folds. An upcoming PR will handle real auto indentation extended to support pasting and deletion. * If code folding was disabled both matching delimiter highlights and block id line were also disabled. * ImHex crashed if lines were inserted while semantic highlighting was taking place which can only occur on very large patterns. The detection of task interruption needed to be fixed. * m_changesWereParsed had the same information as m_hasUnevaluatedChanges, so it was removed and the latter was renamed to the more accurate m_hasUnparsedChanges. * Some functions were moved to more sensible files and some remain to be moved to minimize this Pr changes.
Catch up with latest ImHex codebase
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
14 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Updates the ImHex codebase to the latest master version from https://github.com/WerWolv/ImHex to address a bug in imgui which has been recently resolved in both imgui and ImHex.
This PR simply merges in the latest master changes from
WerWolv/ImHex- there were no conflicts to resolve or anything, so to avoid a lengthy review you could just merge the changes intodavidgiven/ImHexyourself and disregard this PR.See the companion PR for fixes to incompatibilities between fluxengine's Gui2 code and the current ImHex version.