Completion changes, &c., in light of libripgrep

ci/test_complete.sh

@@ -39,12 +39,14 @@ main() {
     print -rl - 'Comparing options:' "-$rg" "+$_rg"
 
     # 'Parse' options out of the `--help` output. To prevent false positives we
-    # only look at lines where the first non-white-space character is `-`
+    # only look at lines where the first non-white-space character is `-`, or
+    # where a long option starting with certain letters (see `_rg`) is found.
+    # Occasionally we may have to handle some manually, however
     help_args=( ${(f)"$(
         $rg --help |
-        $rg -- '^\s*-' |
-        $rg -io -- '[\t ,](-[a-z0-9]|--[a-z0-9-]+)\b' |
-        tr -d '\t ,' |
+        $rg -i -- '^\s+--?[a-z0-9]|--[imnp]' |
+        $rg -ior '$1' -- $'[\t /\"\'`.,](-[a-z0-9]|--[a-z0-9-]+)\\b' |
+        $rg -v -- --print0 | # False positives
         sort -u
     )"} )
 
@@ -58,8 +60,6 @@ main() {
     comp_args=( ${comp_args%%-[:[]*}  ) # Strip everything after -optname-
     comp_args=( ${comp_args%%[:+=[]*} ) # Strip everything after other optspecs
     comp_args=( ${comp_args##[^-]*}   ) # Remove non-options
-
-    # This probably isn't necessary, but we should ensure the same order
     comp_args=( ${(f)"$( print -rl - $comp_args | sort -u )"} )
 
     (( $#help_args )) || {

complete/_rg

@@ -26,8 +26,10 @@ _rg() {
   # style set. Note that this prefix check has to be updated manually to account
   # for all of the potential negation options listed below!
   if
-    # (--[imn]* => --ignore*, --messages, --no-*)
-    [[ $PREFIX$SUFFIX == --[imn]* ]] ||
+    # We also want to list all of these options during testing
+    [[ $_RG_COMPLETE_LIST_ARGS == (1|t*|y*) ]] ||
+    # (--[imnp]* => --ignore*, --messages, --no-*, --pcre2-unicode)
+    [[ $PREFIX$SUFFIX == --[imnp]* ]] ||
     zstyle -t ":complete:$curcontext:*" complete-all
   then
     no=
@@ -61,8 +63,12 @@ _rg() {
     $no"--no-column[don't show column numbers for matches]"
 
     + '(count)' # Counting options
-    '(passthru)'{-c,--count}'[only show count of matching lines for each file]'
-    '(passthru)--count-matches[only show count of individual matches for each file]'
+    {-c,--count}'[only show count of matching lines for each file]'
+    '--count-matches[only show count of individual matches for each file]'
+
+    + '(encoding)' # Encoding options
+    {-E+,--encoding=}'[specify text encoding of files to search]: :_rg_encodings'
+    $no'--no-encoding[use default text encoding]'
 
     + file # File-input options
     '*'{-f+,--file=}'[specify file containing patterns to search for]: :_files'
@@ -111,17 +117,18 @@ _rg() {
     "--no-ignore-vcs[don't respect version control ignore files]"
     $no'--ignore-vcs[respect version control ignore files]'
 
-    + '(json)' # json options
-    '--json[output results in a JSON Lines format]'
-    $no"--no-json[output results in the standard format]"
+    + '(json)' # JSON options
+    '--json[output results in JSON Lines format]'
+    $no"--no-json[don't output results in JSON Lines format]"
 
-    + '(line)' # Line-number options
+    + '(line-number)' # Line-number options
     {-n,--line-number}'[show line numbers for matches]'
     {-N,--no-line-number}"[don't show line numbers for matches]"
 
-    + '(line terminator)' # line terminator options
-    '--crlf[use CRLF as a line terminator]'
-    $no"--no-crlf[do not use CRLF as a line terminator]"
+    + '(line-terminator)' # Line-terminator options
+    '--crlf[use CRLF as line terminator]'
+    $no"--no-crlf[don't use CRLF as line terminator]"
+    '(text)--null-data[use NUL as line terminator]'
 
     + '(max-depth)' # Directory-depth options
     '--max-depth=[specify max number of directories to descend]:number of directories'
@@ -139,23 +146,28 @@ _rg() {
     '--mmap[search using memory maps when possible]'
     "--no-mmap[don't search using memory maps]"
 
-    + '(multiline)' # multiline options
+    + '(multiline)' # Multiline options
     {-U,--multiline}'[permit matching across multiple lines]'
-    $no"--no-multiline[restrict matches to at most one line each]"
-    '--multiline-dotall[make "." match newline in multiline mode]'
+    $no'(multiline-dotall)--no-multiline[restrict matches to at most one line each]'
+
+    + '(multiline-dotall)' # Multiline DOTALL options
+    '(--no-multiline)--multiline-dotall[allow "." to match newline (with -U)]'
+    $no"(--no-multiline)--no-multiline-dotall[don't allow \".\" to match newline (with -U)]"
 
     + '(only)' # Only-match options
-    '(passthru replace)'{-o,--only-matching}'[show only matching part of each line]'
+    {-o,--only-matching}'[show only matching part of each line]'
 
     + '(passthru)' # Pass-through options
-    '(--vimgrep count only replace)--passthru[show both matching and non-matching lines]'
-    '!(--vimgrep count only replace)--passthrough'
+    '(--vimgrep)--passthru[show both matching and non-matching lines]'
+    '!(--vimgrep)--passthrough'
 
     + '(pcre2)' # PCRE2 options
-    {-P,--pcre2}'[Enable matching with PCRE2]'
-    $no"--no-pcre2[don't use PCRE2]"
-    "--pcre2-unicode[Enable PCRE2 Unicode mode]"
-    $no"--pcre2-unicode[Disable PCRE2 Unicode mode]"
+    {-P,--pcre2}'[enable matching with PCRE2]'
+    $no'(pcre2-unicode)--no-pcre2[disable matching with PCRE2]'
+
+    + '(pcre2-unicode)' # PCRE2 Unicode options
+    $no'(--no-pcre2-unicode)--pcre2-unicode[enable PCRE2 Unicode mode (with -P)]'
+    '(--no-pcre2-unicode)--no-pcre2-unicode[disable PCRE2 Unicode mode (with -P)]'
 
     + '(pre)' # Preprocessing options
     '(-z --search-zip)--pre=[specify preprocessor utility]:preprocessor utility:_command_names -e'
@@ -169,24 +181,24 @@ _rg() {
     '(1 file)*'{-e+,--regexp=}'[specify pattern]:pattern'
 
     + '(replace)' # Replacement options
-    '(count only passthru)'{-r+,--replace=}'[specify string used to replace matches]:replace string'
+    {-r+,--replace=}'[specify string used to replace matches]:replace string'
 
     + '(sort)' # File-sorting options
     '(threads)--sort-files[sort results by file path (disables parallelism)]'
     $no"--no-sort-files[don't sort results by file path]"
 
-    + stats # Statistics options
+    + '(stats)' # Statistics options
     '(--files file-match)--stats[show search statistics]'
     $no"--no-stats[don't show search statistics]"
 
     + '(text)' # Binary-search options
     {-a,--text}'[search binary files as if they were text]'
-    $no"--no-text[don't search binary files as if they were text]"
+    $no"(--null-data)--no-text[don't search binary files as if they were text]"
 
     + '(threads)' # Thread-count options
     '(--sort-files)'{-j+,--threads=}'[specify approximate number of threads to use]:number of threads'
 
-    + '(trim)' # trim options
+    + '(trim)' # Trim options
     '--trim[trim any ASCII whitespace prefix from each line]'
     $no"--no-trim[don't trim ASCII whitespace prefix from each line]"
 
@@ -218,7 +230,6 @@ _rg() {
     '--context-separator=[specify string used to separate non-continuous context lines in output]:separator'
     '--debug[show debug messages]'
     '--dfa-size-limit=[specify upper size limit of generated DFA]:DFA size (bytes)'
-    '(-E --encoding)'{-E+,--encoding=}'[specify text encoding of files to search]: :_rg_encodings'
     "(1 stats)--files[show each file that would be searched (but don't search)]"
     '*--ignore-file=[specify additional ignore file]:ignore file:_files'
     '(-v --invert-match)'{-v,--invert-match}'[invert matching]'
@@ -227,7 +238,6 @@ _rg() {
     '--max-filesize=[specify size above which files should be ignored]:file size (bytes)'
     "--no-config[don't load configuration files]"
     '(-0 --null)'{-0,--null}'[print NUL byte after file names]'
-    '--null-data[use NUL as a line terminator]'
     '--path-separator=[specify path separator to use when printing file names]:separator'
     '(-q --quiet)'{-q,--quiet}'[suppress normal output]'
     '--regex-size-limit=[specify upper size limit of compiled regex]:regex size (bytes)'

src/app.rs

@@ -1424,10 +1424,18 @@ This flag is generally intended to be used in an alias or your ripgrep config
 file if you prefer \"dot all\" semantics by default. Note that regardless of
 whether this flag is used, \"dot all\" semantics can still be controlled via
 inline flags in the regex pattern itself, e.g., '(?s:.)' always enables \"dot
-all\" where as '(?-s:.)' always disables \"dot all\".
+all\" whereas '(?-s:.)' always disables \"dot all\".
+
+This flag can be disabled with --no-multiline-dotall.
 ");
     let arg = RGArg::switch("multiline-dotall")
-        .help(SHORT).long_help(LONG);
+        .help(SHORT).long_help(LONG)
+        .overrides("no-multiline-dotall");
+    args.push(arg);
+
+    let arg = RGArg::switch("no-multiline-dotall")
+        .hidden()
+        .overrides("multiline-dotall");
     args.push(arg);
 }
 
@@ -1469,7 +1477,7 @@ fn flag_no_ignore_global(args: &mut Vec<RGArg>) {
     const LONG: &str = long!("\
 Don't respect ignore files that come from \"global\" sources such as git's
 `core.excludesFile` configuration option (which defaults to
-`$HOME/.config/git/ignore).
+`$HOME/.config/git/ignore`).
 
 This flag can be disabled with the --ignore-global flag.
 ");
@@ -1581,11 +1589,11 @@ the default of '\\n'.
 
 This is useful when searching large binary files that would otherwise have very
 long lines if '\\n' were used as the line terminator. In particular, ripgrep
-requires that, at a minimum, each line must fit into memory. Use NUL instead
+requires that, at a minimum, each line must fit into memory. Using NUL instead
 can be a useful stopgap to keep memory requirements low and avoid OOM (out of
 memory) conditions.
 
-This is also useful for processing NUL delimited data, such that that emitted
+This is also useful for processing NUL delimited data, such as that emitted
 when using ripgrep's -0/--null flag or find's --print0 flag.
 
 Using this flag implies -a/--text.
@@ -1684,7 +1692,7 @@ enabled. Instead, PCRE2 *requires* that everything it searches when Unicode
 mode is enabled is valid UTF-8. (Or valid UTF-16/UTF-32, but for the purposes
 of ripgrep, we only discuss UTF-8.) This means that if you have PCRE2's Unicode
 mode enabled and you attempt to search invalid UTF-8, then the search for that
-file will hault and print an error. For this reason, when PCRE2's Unicode mode
+file will halt and print an error. For this reason, when PCRE2's Unicode mode
 is enabled, ripgrep will automatically \"fix\" invalid UTF-8 sequences by
 replacing them with the Unicode replacement codepoint.