This commit hacks in a bug fix for handling look-around across multiple
lines. The main problem is that by the time the matching lines are sent
to the printer, the surrounding context---which some look-behind or
look-ahead might have matched---could have been dropped if it wasn't
part of the set of matching lines. Therefore, when the printer re-runs
the regex engine in some cases (to do replacements, color matches, etc
etc), it won't be guaranteed to see the same matches that the searcher
found.
Overall, this is a giant clusterfuck and suggests that the way I divided
the abstraction boundary between the printer and the searcher is just
wrong. It's likely that the searcher needs to handle more of the work of
matching and pass that info on to the printer. The tricky part is that
this additional work isn't always needed. Ultimately, this means a
serious re-design of the interface between searching and printing. Sigh.
The way this fix works is to smuggle the underlying buffer used by the
searcher through into the printer. Since these bugs only impact
multi-line search (otherwise, searches are only limited to matches
across a single line), and since multi-line search always requires
having the entire file contents in a single contiguous slice (memory
mapped or on the heap), it follows that the buffer we pass through when
we need it is, in fact, the entire haystack. So this commit refactors
the printer's regex searching to use that buffer instead of the intended
bundle of bytes containing just the relevant matching portions of that
same buffer.
There is one last little hiccup: PCRE2 doesn't seem to have a way to
specify an ending position for a search. So when we re-run the search to
find matches, we can't say, "but don't search past here." Since the
buffer is likely to contain the entire file, we really cannot do
anything here other than specify a fixed upper bound on the number of
bytes to search. So if look-ahead goes more than N bytes beyond the
match, this code will break by simply being unable to find the match. In
practice, this is probably pretty rare. I believe that if we did a
better fix for this bug by fixing the interfaces, then we'd probably try
to have PCRE2 find the pertinent matches up front so that it never needs
to re-discover them.
Fixes#1412
This commit fixes a subtle bug in multi-line replacement of line
terminators.
The problem is that even though ripgrep supports multi-line searches, it
is *still* line oriented. It still needs to print line numbers, for
example. For this reason, there are various parts in the printer that
iterate over lines in order to format them into the desired output.
This turns out to be problematic in some cases. #1311 documents one of
those cases (with line numbers enabled to highlight a point later):
$ printf "hello\nworld\n" | rg -n -U "\n" -r "?"
1:hello?
2:world?
But the desired output is this:
$ printf "hello\nworld\n" | rg -n -U "\n" -r "?"
1:hello?world?
At first I had thought that the main problem was that the printer was
taking ownership of writing line terminators, even if the input already
had them. But it's more subtle than that. If we fix that issue, we get
output like this instead:
$ printf "hello\nworld\n" | rg -n -U "\n" -r "?"
1:hello?2:world?
Notice how '2:' is printed before 'world?'. The reason it works this way
is because matches are reported to the printer in a line oriented way.
That is, the printer gets a block of lines. The searcher guarantees that
all matches that start or end in any of those lines also end or start in
another line in that same block. As a result, the printer uses this
assumption: once it has processed a block of lines, the next match will
begin on a new and distinct line. Thus, things like '2:' are printed.
This is generally all fine and good, but an impedance mismatch arises
when replacements are used. Because now, the replacement can be used to
change the "block of lines" approach. Now, in terms of the output, the
subsequent match might actually continue the current line since the
replacement might get rid of the concept of lines altogether.
We can sometimes work around this. For example:
$ printf "hello\nworld\n" | rg -U "\n(.)?" -r '?$1'
hello?world?
Why does this work? It's because the '(.)' after the '\n' causes the
match to overlap between lines. Thus, the searcher guarantees that the
block sent to the printer contains every line.
And there in lay the solution: all we need to do is tweak the multi-line
searcher so that it combines lines with matches that directly adjacent,
instead of requiring at least one byte of overlap. Fixing that solves
the issue above. It does cause some tests to fail:
* The binary3 test in the searcher crate fails because adjacent line
matches are now one part of block, and that block is scanned for
binary data. To preserve the essence of the test, we insert a couple
dummy lines to split up the blocks.
* The JSON CRLF test. It was testing that we didn't output any messages
with an empty 'submatches' array. That is indeed still the case. The
difference is that the messages got combined because of the adjacent
line merging behavior. This is a slight change to the output, but is
still correct.
Fixes#1311
It turns out that the vimgrep format really only wants one line per
match, even when that match spans multiple lines.
We continue to support the previous behavior (print all lines in a
match) in the `grep-printer` crate. We add a new option to enable the
"only print the first line" behavior, and unconditionally enable it in
ripgrep. We can do that because the option has no effect in single-line
mode, since, well, in that case matches are guaranteed to span one line
anyway.
Fixes#1866
These flags permit configuring the bytes used to delimit fields in match
or context lines, where "fields" are things like the file path, line
number, column number and the match/context itself.
Fixes#1842, Closes#1871
Previous, 'foo/**' would match 'foo', but it shouldn't have. In this
case, not matching 'foo' is what is documented and also seems consistent
with other recursive globbing implementations (like that in zsh).
This also updates the prefix extractor to pull 'foo/' out of 'foo/**'.
Closes#1756
This seems like an obvious optimization but becomes critical when
filesystem operations even as simple as stat can result in significant
overheads; an example of this was a bespoke filesystem layer in Windows
that hosted files remotely and would download them on-demand when
particular filesystem operations occurred. Users of this system who
ensured correct file-type fileters were being used could still get
unnecessary file access resulting in large downloads.
Fixes#1657, Closes#1660
In the case where after-context is requested with a match count limit,
we need to be careful not to reset the state tracking the remaining
context lines.
Fixes#1380, Closes#1642
Previously, we were only looking for the UTF-16 BOM for determining
whether to do transcoding or not. But we should also look for the UTF-8
BOM as well.
Fixes#1638, Closes#1697
This is somewhat non-standard, but it seems nice on the surface: short
flag names are in short supply, --hidden is probably somewhat common and
-. has an obvious connection with how hidden files are named on Unix.
Closes#1680
This fixes a bug where it was assumed that 'is_suffix' when CRLF
handling was enabled mean that '\r\n' was present. But that's not the
case, and it is intentional that 'is_suffix' only looks for '\n'. (Which
is why #1803 wasn't taken, which tries to fix this by changing
'is_suffix'.)
Fixes#1765, Closes#1803
This was once part of ripgrep, but at some point, was unintentionally
removed. The value of this warning is that since ripgrep tries to be
"smart" by default, it can be surprising if it doesn't search certain
things. This warning covers the case when ripgrep searches *nothing*,
which happens somewhat more frequently than you might expect. e.g., If
you're searching within an ignore directory.
Note that for now, we only print this message when the user has not
supplied any explicit paths. It's not clear that we want to print this
otherwise, and in particular, it seems that the message shows up too
eagerly. e.g., 'rg foo does-not-exist' will both print an error about
'does-not-exist' not existing, *and* the message about no files being
searched, which seems annoying in this case. We can always refine this
logic later.
Fixes#1404, Closes#1762
This also replaces '--all' in Cargo commands with '--workspace'. The
former has apparently been deprecated.
We also fix a couple warnings that this new step detected.
Closes#1848
If ripgrep was called in a way where the entire contents of a file
aren't read (like --files-with-matches, among other methods), and if the
file was read through an external process, then ripgrep would never reap
that process.
We fix this by introducing an explicit 'close' method, which we now call
when using decompression or preprocessor searches.
The implementation of 'close' is a little hokey. In particular, when we
close stdout, this usually results in a broken pipe, and, consequently,
a non-zero code returned once the child process is reaped. This is
"situation normal," so we invent a (hopefully portable) heuristic for
detecting it.
Fixes#1766, Closes#1767
This fixes a bug only present on Windows that would permit someone to
execute an arbitrary program if they crafted an appropriate directory
tree. Namely, if someone put an executable named 'xz.exe' in the root of
a directory tree and one ran 'rg -z foo' from the root of that tree,
then the 'xz.exe' executable in that tree would execute if there are any
'xz' files anywhere in the tree.
The root cause of this problem is that 'CreateProcess' on Windows will
implicitly look in the current working directory for an executable when
it is given a relative path to a program. Rust's standard library allows
this behavior to occur, so we work around it here. We work around it by
explicitly resolving programs like 'xz' via 'PATH'. That way, we only
ever pass an absolute path to 'CreateProcess', which avoids the implicit
behavior of checking the current working directory.
This fix doesn't apply to non-Windows systems as it is believed to only
impact Windows. In theory, the bug could apply on Unix if '.' is in
one's PATH, but at that point, you reap what you sow.
While the extent to which this is a security problem isn't clear, I
think users generally expect to be able to download or clone
repositories from the Internet and run ripgrep on them without fear of
anything too awful happening. Being able to execute an arbitrary program
probably violates that expectation. Therefore, CVE-2021-3013[1] was
created for this issue.
We apply the same logic to the --pre command, since the --pre command is
likely in a user's config file and it would be surprising for something
that the user is searching to modify which preprocessor command is used.
The --pre and -z/--search-zip flags are the only two ways that ripgrep
will invoke external programs, so this should cover any possible
exploitable cases of this bug.
[1] - https://cve.mitre.org/cgi-bin/cvename.cgi?name=CVE-2021-3013
This fixes a bug where using \A or (?-m)^ in combination with
-U/--multiline would permit matches that aren't anchored to the
beginning of the file. The underlying cause was an optimization that
occurred when mmaps couldn't be used. Namely, ripgrep tries to still
read the input incrementally if it knows the pattern can't match through
a new line. But the detection logic was flawed, since it didn't account
for line anchors. This commit fixes that.
Fixes#1878, Fixes#1879
It turned out that --vimgrep wasn't quite getting the column of each
match correctly. Instead of printing column numbers relative to the
current line, it was printing column numbers as byte offsets relative to
where the match began. To fix this, we simply subtract the offset of the
line number from the beginning of the match. If the beginning of the
match came before the start of the current line, then there's really
nothing sensible we can do other than to use a column number of 1, which
we now document.
Interestingly, existing tests were checking that the previous behavior
was intended. My only defense is that I somehow tricked myself into
thinking it was a byte offset instead of a column number.
Kudos to @bfrg for calling this out in #1866:
https://github.com/BurntSushi/ripgrep/issues/1866#issuecomment-841635553
It looks like no new dependencies have been introduced. Yay!
This update was primarily motivated to bring regex 1.5 in with its new
memmem implementation from the memchr crate.
This message will emit the binary detection mechanism being used for
each file.
This does not noticeably increases the number of log messages, as the
'trace' level is already used for emitting messages for every file
searched.
This trace message was added in the course of investigating #1838.
Sadly, there were several tests that are coupled to the size of the
buffer used by ripgrep. Making the tests agnostic to the size is
difficult. And it's annoying to fix the tests. But we rarely change the
buffer size, so ¯\_(ツ)_/¯.
This increases the initial buffer size from 8KB to 64KB. This actually
leads to a reasonably noticeable improvement in at least one work-load,
and is unlikely to regress in any other case. Also, since Rust programs
(at least on Linux) seem to always use a minimum of 6-8MB of memory,
adding an extra 56KB is negligible.
Before:
$ hyperfine -i "rg 'zqzqzqzq' OpenSubtitles2018.raw.en --no-mmap"
Benchmark #1: rg 'zqzqzqzq' OpenSubtitles2018.raw.en --no-mmap
Time (mean ± σ): 2.109 s ± 0.012 s [User: 565.5 ms, System: 1541.6 ms]
Range (min … max): 2.094 s … 2.128 s 10 runs
After:
$ hyperfine -i "rg 'zqzqzqzq' OpenSubtitles2018.raw.en --no-mmap"
Benchmark #1: rg 'zqzqzqzq' OpenSubtitles2018.raw.en --no-mmap
Time (mean ± σ): 1.802 s ± 0.006 s [User: 462.3 ms, System: 1337.9 ms]
Range (min … max): 1.795 s … 1.814 s 10 runs
If ripgrep hasn't been built yet (i.e. target/debug/ doesn't exist),
then cargo-out-dir can't find OUT_DIR and the copy commands fail. Fix by
running cargo build before finding OUT_DIR.
Also add a check to fail early with a sensible error message when
asciidoctor isn't installed, rather than failing because of a missing
rg.1 file after the build.
PR #1831
memmap is unmaintained at this point and it is being flagged as a
RUSTSEC advisory in ripgrep. This doesn't seem like that big of a deal
to me honestly, but memmap2 looks like a fine choice at this point.
Fixes#1785, Closes#1786
It seems that PowerShell uses sockets instead of FIFOs to redirect the
output between commands. So add `is_socket` to our `is_readable_stdin`
check.
This seems unlikely to cause problems and it probably more generally
correct than what we had before. In theory, it could cause problems if
it produces false positives, in which case, ripgrep will try to read
stdin when it should search the current working directory. (And this
usually winds up manifesting as ripgrep blocking forever.) But, if the
stdin handle reports itself as a socket, then it seems like we should
read it.
Fixes#1741, Closes#1742
`ignore::Error` wraps `std::io::Error` with additional information
(as well as expose non-IO errors). For people wanting to inspect what
the error is, they have to recursively match the Enum. This provides
`io_error` and `into_io_error` helpers to do this for the user.
PR #1740
