QFindDoc.tioga
Richard E. Sweet, December 5, 1985 10:51:45 am PST
Eric Nickell, March 25, 1987 10:44:52 pm PST
Bloomenthal, October 8, 1990 12:06 pm PDT
Michael Plass, February 20, 1992 1:37 pm PST
Last tweaked by Mike Spreitzer February 7, 1992 6:05 pm PST
QFIND
CEDAR 10.0 FOR INTERNAL XEROX USE ONLY
QFind
-- a quick simple file search program
Richard E. Sweet
© Copyright 1987, 1990, 1992 Xerox Corporation. All rights reserved.
Abstract: QFind is a file scanning program similar in spirit to Grep. It lacks most of the functionality of Grep, but makes up for it by being 3-4 times as fast.
Created by: Richard E. Sweet
Maintained by: Michael Plass:PARC:Xerox
Keywords: searching
XEROX Xerox Corporation
Palo Alto Research Center
3333 Coyote Hill Road
Palo Alto, California 94304
For Internal Xerox Use Only
QFind is a file scanning program similar in spirit to Grep. It lacks most of the functionality of Grep, but makes up for this by being 3-4 times as fast. It finds only simple text keys (no regular expressions, wild cards, etc.).
QFind has the following optional switches, which must immediately follow the command name:
-c be case sensitive
-f list file names only
-o open files successfully searched
-d <DF file name> search the files in the Directory and Exports clauses of the named DF file
-m use source version maps to supply directory information
-p print file position of first match along with name (useful with -f only)
-q the last option; use if your search key begins with a dash
-r <DF file name> DF/module search
-s print the short names of the files and separate the file listings
-w match words only
Typical usages:
QFind <search key> <list of files>
QFind -c <search key> <list of files>
QFind -c -f -m <search key> <list of short names>
If needed, a ".mesa" extension is supplied before version-map lookup
QFind -fcm <search key> <list of short names>
Same as above; switches without arguments may be combined for brevity
QFind -d <DF file name> <search key>
QFind -d <DF file name> <search key> <list of files>
Searches the files in the list, plus those in the Directory clauses of the named DF file
QFind -wc -r [PCedar2.0]<Top>PCedar.df ViewersWorldExtras.CallWhenWorldSet
Searches for uses of ViewersWorldExtras.CallWhenWorldSet in the PCedar release.
QFind -wc CallWhenWorldSet $(CitingFiles ViewersWorldExtras)
Searches for uses of ViewersWorldExtras.CallWhenWorldSet in the PCedar release --- much more efficiently, since /r/PCedar.depends is consulted to find the list of candidate files.
QFind -wc -r $ ViewersWorldExtras.CallWhenWorldSet $(CitingFiles ViewersWorldExtras)
Like above, but tests in a slightly more specific way (the file must contain the token "ViewersWorldExtras" as well as "CallWhenWorldSet"). This command assumes there's no DF "$".
If <search key> contains blanks, it must be enclosed in quotes ("). This is approximately equivalent to Grep -o [-c] [-w] <search key> <list of files>. The number of matches reported by the two programs differs in that QFind, after printing out the line containing a match, doesn't bother to look on the rest of that line for further matches.
A DF/module search is useful for answering the question: who references item X from interface Y? The key must be of the form Y.X. The DF files given with the -r switch are elaborated. Whenever a suite either defines or imports Y.mob, the source files in that suite are considered. The source files resulting from all command switches are filtered according to whether they include the token Y; those that pass are searched for X. A suite is the DFs (excluding the subsidiary suites) included in a DF whose base name either ends in "-Suite" or lacks a dash. The base name of a file name is the part of the last component before its extension. For example, [PCedar2.0]<Top>PCedar.df heads a suite, which includes a bunch of other DFs that head their own suites; there is only one DF (PCedar.df itself) in the suite headed by PCedar.df.
QFind uses the Boyer-Moore sub-linear search algorithm; the longer the key, the faster the search.