Uh oh!
There was an error while loading. Please reload this page.
Support process substitution for --filelist= - #4349
Merged
Merged
Conversation
Cyan4973
marked this pull request as draft
March 25, 2025 22:13
Cyan4973
marked this pull request as ready for review
March 25, 2025 23:40
and more narrow scope of local variables
jaypatrickhoward added a commit
to jaypatrickhoward/zstd
that referenced
this pull request
Sep 7, 2026
`--filelist` stopped accepting Windows CRLF line endings in 165e52c ("first implementation supporting Process Substitution", facebook#4349). That rewrite reads the list into a buffer opened in binary mode and splits it on '\n' alone. Previously the list was opened in text mode, where the Windows CRT translated CRLF to LF before the parser saw it, so stripping '\n' was sufficient. In binary mode the CR survives into every path: zstd: can't stat a.txt : No such file or directory -- ignored The '\r' is invisible in that message, which makes it awkward to diagnose. Strip a trailing '\r' when the line pointers are built, but only on Windows, where CRLF is the native line separator and a '\r' cannot be used in a filename. Elsewhere '\r' is a legal filename byte and is left alone, so behaviour on those platforms is unchanged. This restores the pre-regression behaviour exactly, on every platform, without adding a new one. Adds a playTests case, guarded to Windows, that fails without the fix.
jaypatrickhoward added a commit
to jaypatrickhoward/zstd
that referenced
this pull request
Sep 7, 2026
`--filelist` stopped accepting Windows CRLF line endings in 165e52c ("first implementation supporting Process Substitution", facebook#4349). That rewrite reads the list into a buffer opened in binary mode and splits it on '\n' alone. Previously the list was opened in text mode, where the Windows CRT translated CRLF to LF before the parser saw it, so stripping '\n' was sufficient. In binary mode the CR survives into every path: zstd: can't stat a.txt : No such file or directory -- ignored The '\r' is invisible in that message, which makes it awkward to diagnose. Strip a trailing '\r' when the line pointers are built, but only on Windows, where CRLF is the native line separator and a '\r' cannot be used in a filename. Elsewhere '\r' is a legal filename byte and is left alone, so behaviour on those platforms is unchanged. This restores the pre-regression behaviour exactly, on every platform, without adding a new one. Adds a playTests case, guarded to Windows, that fails without the fix.
jaypatrickhoward added a commit
to jaypatrickhoward/zstd
that referenced
this pull request
Sep 7, 2026
`--filelist` stopped accepting Windows CRLF line endings in 165e52c ("first implementation supporting Process Substitution", facebook#4349). That rewrite reads the list into a buffer opened in binary mode and splits it on '\n' alone. Previously the list was opened in text mode, where the Windows CRT translated CRLF to LF before the parser saw it, so stripping '\n' was sufficient. In binary mode the CR survives into every path: zstd: can't stat a.txt : No such file or directory -- ignored The '\r' is invisible in that message, which makes it awkward to diagnose. Strip a trailing '\r' when the line pointers are built, but only on Windows, where CRLF is the native line separator and a '\r' cannot be used in a filename. Elsewhere '\r' is a legal filename byte and is left alone, so behaviour on those platforms is unchanged. This restores the pre-regression behaviour exactly, on every platform, without adding a new one. Adds a playTests case, guarded to Windows, that fails without the fix.
jaypatrickhoward added a commit
to jaypatrickhoward/zstd
that referenced
this pull request
Sep 7, 2026
`--filelist` stopped accepting Windows CRLF line endings in 165e52c ("first implementation supporting Process Substitution", facebook#4349). That rewrite reads the list into a buffer opened in binary mode and splits it on '\n' alone. Previously the list was opened in text mode, where the Windows CRT translated CRLF to LF before the parser saw it, so stripping '\n' was sufficient. In binary mode the CR survives into every path: zstd: can't stat a.txt : No such file or directory -- ignored The '\r' is invisible in that message, which makes it awkward to diagnose. Strip a trailing '\r' when the line pointers are built, but only on Windows, where CRLF is the native line separator and a '\r' cannot be used in a filename. Elsewhere '\r' is a legal filename byte and is left alone, so behaviour on those platforms is unchanged. This restores the pre-regression behaviour exactly, on every platform, without adding a new one. Adds a playTests case, guarded to Windows, that fails without the fix.
jaypatrickhoward added a commit
to jaypatrickhoward/zstd
that referenced
this pull request
Sep 7, 2026
`--filelist` stopped accepting Windows CRLF line endings in 165e52c ("first implementation supporting Process Substitution", facebook#4349). That rewrite reads the list into a buffer opened in binary mode and splits it on '\n' alone. Previously the list was opened in text mode, where the Windows CRT translated CRLF to LF before the parser saw it, so stripping '\n' was sufficient. In binary mode the CR survives into every path: zstd: can't stat a.txt : No such file or directory -- ignored The '\r' is invisible in that message, which makes it awkward to diagnose. Strip a trailing '\r' when the line pointers are built, but only on Windows, where CRLF is the native line separator and a '\r' cannot be used in a filename. Elsewhere '\r' is a legal filename byte and is left alone, so behaviour on those platforms is unchanged. This restores the pre-regression behaviour exactly, on every platform, without adding a new one. Adds a playTests case, guarded to Windows, that fails without the fix.
jaypatrickhoward added a commit
to jaypatrickhoward/zstd
that referenced
this pull request
Sep 7, 2026
`--filelist` stopped accepting Windows CRLF line endings in 165e52c ("first implementation supporting Process Substitution", facebook#4349). That rewrite reads the list into a buffer opened in binary mode and splits it on '\n' alone. Previously the list was opened in text mode, where the Windows CRT translated CRLF to LF before the parser saw it, so stripping '\n' was sufficient. In binary mode the CR survives into every path: zstd: can't stat a.txt : No such file or directory -- ignored The '\r' is invisible in that message, which makes it awkward to diagnose. Strip a trailing '\r' when the line pointers are built, but only on Windows, where CRLF is the native line separator and a '\r' cannot be used in a filename. Elsewhere '\r' is a legal filename byte and is left alone, so behaviour on those platforms is unchanged. This restores the pre-regression behaviour exactly, on every platform, without adding a new one. Adds a playTests case, guarded to Windows, that fails without the fix.
jaypatrickhoward added a commit
to jaypatrickhoward/zstd
that referenced
this pull request
Sep 7, 2026
`--filelist` stopped accepting Windows CRLF line endings in 165e52c ("first implementation supporting Process Substitution", facebook#4349). That rewrite reads the list into a buffer opened in binary mode and splits it on '\n' alone. Previously the list was opened in text mode, where the Windows CRT translated CRLF to LF before the parser saw it, so stripping '\n' was sufficient. In binary mode the CR survives into every path: zstd: can't stat a.txt : No such file or directory -- ignored The '\r' is invisible in that message, which makes it awkward to diagnose. Strip a trailing '\r' when the line pointers are built, but only on Windows, where CRLF is the native line separator and a '\r' cannot be used in a filename. Elsewhere '\r' is a legal filename byte and is left alone, so behaviour on those platforms is unchanged. This restores the pre-regression behaviour exactly, on every platform, without adding a new one. Adds a playTests case, guarded to Windows, that fails without the fix.
jaypatrickhoward added a commit
to jaypatrickhoward/zstd
that referenced
this pull request
Sep 7, 2026
`--filelist` stopped accepting Windows CRLF line endings in 165e52c ("first implementation supporting Process Substitution", facebook#4349). That rewrite reads the list into a buffer opened in binary mode and splits it on '\n' alone. Previously the list was opened in text mode, where the Windows CRT translated CRLF to LF before the parser saw it, so stripping '\n' was sufficient. In binary mode the CR survives into every path: zstd: can't stat a.txt : No such file or directory -- ignored The '\r' is invisible in that message, which makes it awkward to diagnose. Strip a trailing '\r' when the line pointers are built, but only on Windows, where CRLF is the native line separator and a '\r' cannot be used in a filename. Elsewhere '\r' is a legal filename byte and is left alone, so behaviour on those platforms is unchanged. This restores the pre-regression behaviour exactly, on every platform, without adding a new one. Adds a playTests case, guarded to Windows, that fails without the fix.
jaypatrickhoward added a commit
to jaypatrickhoward/zstd
that referenced
this pull request
Sep 7, 2026
`--filelist` stopped accepting Windows CRLF line endings in 165e52c ("first implementation supporting Process Substitution", facebook#4349). That rewrite reads the list into a buffer opened in binary mode and splits it on '\n' alone. Previously the list was opened in text mode, where the Windows CRT translated CRLF to LF before the parser saw it, so stripping '\n' was sufficient. In binary mode the CR survives into every path: zstd: can't stat a.txt : No such file or directory -- ignored The '\r' is invisible in that message, which makes it awkward to diagnose. Strip a trailing '\r' when the line pointers are built, but only on Windows, where CRLF is the native line separator and a '\r' cannot be used in a filename. Elsewhere '\r' is a legal filename byte and is left alone, so behaviour on those platforms is unchanged. This restores the pre-regression behaviour exactly, on every platform, without adding a new one. Adds a playTests case, guarded to Windows, that fails without the fix.
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 freeto join this conversation on GitHub.
Already have an account?
Sign in to comment
2 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.
--filelistcapability was designed with the assumption that the source is a file, so it would be possible to know its size in advance, for allocation.Rewrote the capability, so that it can stream data, and discover source size at the end.
solves #4340
The new capability has been tested locally, on a posix laptop.
An issue is on the testing side: process substitution is not supported by
sh, and all our shell test scripts are usingshso far, for broader compatibility with diverse clients.Adding a
bashscript is not in itself a problem, but then it can not be blindly started on any platform, since some do not supportbash. Hence the corresponding CI test must be explicitly triggered.