RECOLLQ(1)

📦
version 1.44.2 · 4 versions
, in arch rolling, epel 10, fedora 43, fedora 44, ubuntu resolute, ubuntu stonking

💖 RECOLLQ 💖 (yes bestie, this is the recollq(1) page)

okay so. gather round. recollq is the command line / standard output Recoll query command, and she is iconic. this is the General Commands Manual entry for her. let's get into it 🎀

✨ SYNOPSIS ✨

recollq [ -c <configdir> ] [ -o | -f | -a ] [ -b ] [ -C ] [ -d ] [ -A ]
[ -p | -q ] [ -e ] [ -m ] [ -n <[first-]cnt> ] [ -Q ] [ -s <stemming
language> ] [ -S <fldname> ] [ -D ] [ -i <additional index directory> ]
[ -F <quoted space separated field name list> ] [ --extract-to <file
path> ] <query string>

recollq -P

🌸 DESCRIPTION 🌸

so the recollq command will execute the Recoll query specified on the command line and print the results to the standard output. that's the whole vibe. she's primarily designed for diagnostics, or for piping the data to some other program 🫶 and like, the basic format and its variations can genuinely be useful for command line querying. no cap.

important bestie moment: the -F option should exclusively be used for using the output data in another program, because it is the only one for which output is guaranteed to be fully parseable. like, the ONLY one. don't be out here parsing the other formats, she will not cooperate 💀

the -c option specifies the configuration directory name, overriding the default or $RECOLL_CONFDIR. so if you want a different config dir, that's your girl.

the query string is built by concatenating all arguments found at the end of the command line (after the options). by default it'll be interpreted as a query language string. quoting should be used as needed to escape characters that might be interpreted by the shell (ie: wildcards) 🐚✨

and now, the little modifiers, one by one:

  • -a — if specified, the query string will be interpreted as an all words simple search query.
  • -o — if specified, the query string will be interpreted as an any word simple search query.
  • -f — if specified, the query string will be interpreted as a file name simple search query.
  • -b — (basic) can be specified to only print the result urls in the output stream. short and sweet, we love her.
  • -d — if set, the text for the result files contents will be dumped to stdout. dumping. the actual file contents. okay 😭
  • -C — with this, the program will only show a single result for multiple identical documents (same data hash). dedupe queen 💅
  • -m — if set, the whole metadata array will be dumped for each document. all of it. every last crumb.
  • -A — if set, the document abstracts will be printed. and with an additional -p (or -q) option, snippets with page numbers, (or line numbers), will be shown instead, if available.
  • -S <fieldname> — sorts the results according to the specified field. use -D for descending order. ⬇️
  • -n <[first-]cnt> — can be used to set the maximum number of results that should be printed. the default is 2000. use a value of 0 for no limit. and if the argument is of the form first-cnt, it also defines the first result to output (from 0). so you get a lil range, cute.
  • -s <language> — selects the word stemming language. the value should match an existing stemming database (as set in the configuration or added with recollindex -s).
  • -i <extra dbdir> — adds the specified Xapian index to the set used for the query. can be specified multiple times. stack 'em 🌈
  • -F <space separated field list> — (one argument, e.g. "author title") should be used for piping the data to another program. after 2 initial lines showing the actual query and the estimated result counts, it will print one line for each result document. each line will have exactly the fields requested on the command line. fields are encoded in base64 and separated by one space character. empty fields are indicated by consecutive space characters. and there is one additional space character at the end of each line. yes, one. it's on purpose. she's precise like that ✨
  • --extract-to <file path> — will extract the first result document of the query to the argument path, which must not exist. use -n first-cnt to select the document. so, like, don't pre-create that file bestie 🙈
  • -P — recollq -P (Period) will print the minimum and maximum modification years for documents in the index. period. 📅
tiny tea: the synopsis also shows -e and -Q floating around in there, and this page never actually says what they do. mystery icons, we respect them from afar 😭

🦋 SEE ALSO 🦋

recollindex(1) recollq(1) recoll.conf(5)

go read those too, they're the rest of the friend group 🫶

and that's the whole page, bestie. this was RECOLLQ(1) from the General Commands Manual, dated 13 November 2007 — an absolute vintage icon, we stan a classic. mwah 💅💖

📦 where this page comes from