You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@subversion.apache.org by Branko Čibej <br...@xbc.nu> on 2001/07/25 21:43:32 UTC
String type madness
I'm totally confused.
Can somebody remind me when we're supposed to use `const svn_string_t*'
vs. `const char*'? In the public interfaces, that is. I seem to remember
GregS saying something about why svn_string_t is good for scripting
language wrappers.
The reason I'm asking is this: I'm totally swamped, working on two
projects simultaneously at work, so I decided to postpone the delta
combiner until after M3 (more on that later), since it's not critical
for this milestone.
Instead, I decided to resurrect a config file parsing library -- which /
should/ be in M3, IMHO -- I have lying around (It's LGPL, but I'm the
author, got a loopohole :-). I'm banging it into shape for Subversion,
and want to get the public interfaces right. Here's an example of what
I'm talking about:
/* Find the value of a (SECTION, OPTION) pair in CFG. Return
DEFAULT_VALUE if the value doesn't exist. */
const char *svn_config_get (svn_config_t *cfg,
const char *section, const char *option,
const char *default_value);
I really see no reason why the parameters should be anything but `const
char*'; but what about the return value? I was considering returning a
svn_string_t instead (IIRC, returning structs is even worse for
scripting wrappers, at least in my experiences with swig), or adding a
svn_string_t out parameter, like this:
void svn_config_get (svn_config_t *cfg, svn_string_t *valuep,
const char *section, const char *option,
const char *default_value);
I'd appreciate some input here.
Brane
P.S.: I plan to tweak my library a bit so that it reads config files
that have the same format as Python's ConfigParser files. I think
interoperability is nice, don't you? Oh, and we'll be able to read the
config straight from the registry on Win32.
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org
Re: String type madness
Posted by Daniel Rall <dl...@finemaltcoding.com>.
Ben Collins-Sussman <su...@collab.net> writes:
> "Paul D. Smith" <pa...@nortelnetworks.com> writes:
>
> > Install viewcvs; it's much better about all these things, including
> > Attic files...
>
> Gee, I wonder who wrote viewcvs? ;-)
Heh. The only reason Tigris has avoided ViewCVS is because it's
written in Python. This is *not* because we have anything against
Python (quite the contrary, actually); it's solely to reduce the
number of language interpretters necessary to run a Tigris
installation. We already have an enormous number of requirements and
technologies. :-[
Thanks, Dan
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org
Re: String type madness
Posted by "Paul D. Smith" <pa...@nortelnetworks.com>.
%% Ben Collins-Sussman <su...@collab.net> writes:
bc> "Paul D. Smith" <pa...@nortelnetworks.com> writes:
>> Install viewcvs; it's much better about all these things, including
>> Attic files...
bc> Gee, I wonder who wrote viewcvs? ;-)
I think Greg Stein wrote it... _I_ certainly had nothing to do with it
:).
But, I did have this exact same problem in my internal SourceForge site,
and solved it by installing viewcvs--works great, less filling.
Does require Python, though... if that matters.
--
-------------------------------------------------------------------------------
Paul D. Smith <ps...@baynetworks.com> HASMAT--HA Software Methods & Tools
"Please remain calm...I may be mad, but I am a professional." --Mad Scientist
-------------------------------------------------------------------------------
These are my opinions---Nortel Networks takes no responsibility for them.
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org
Re: String type madness
Posted by Ben Collins-Sussman <su...@collab.net>.
"Paul D. Smith" <pa...@nortelnetworks.com> writes:
> Install viewcvs; it's much better about all these things, including
> Attic files...
Gee, I wonder who wrote viewcvs? ;-)
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org
Re: String type madness
Posted by "Paul D. Smith" <pa...@nortelnetworks.com>.
Install viewcvs; it's much better about all these things, including
Attic files...
--
-------------------------------------------------------------------------------
Paul D. Smith <ps...@baynetworks.com> HASMAT--HA Software Methods & Tools
"Please remain calm...I may be mad, but I am a professional." --Mad Scientist
-------------------------------------------------------------------------------
These are my opinions---Nortel Networks takes no responsibility for them.
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org
Re: String type madness
Posted by Ben Collins-Sussman <su...@collab.net>.
Silly CVSWeb won't show head revisions if they're deleted. I got
these files by looking at the penultimate revisions.
------------- svn_parse.h
/* svn_parse: shared parsing routines for reading config files
*
* ================================================================
* Copyright (c) 2000 Collab.Net. All rights reserved.
*
* Redistribution and use in source and binary forms, with or without
* modification, are permitted provided that the following conditions are
* met:
*
* 1. Redistributions of source code must retain the above copyright
* notice, this list of conditions and the following disclaimer.
*
* 2. Redistributions in binary form must reproduce the above copyright
* notice, this list of conditions and the following disclaimer in the
* documentation and/or other materials provided with the distribution.
*
* 3. The end-user documentation included with the redistribution, if
* any, must include the following acknowlegement: "This product includes
* software developed by Collab.Net (http://www.Collab.Net/)."
* Alternately, this acknowlegement may appear in the software itself, if
* and wherever such third-party acknowlegements normally appear.
*
* 4. The hosted project names must not be used to endorse or promote
* products derived from this software without prior written
* permission. For written permission, please contact info@collab.net.
*
* 5. Products derived from this software may not use the "Tigris" name
* nor may "Tigris" appear in their names without prior written
* permission of Collab.Net.
*
* THIS SOFTWARE IS PROVIDED ``AS IS'' AND ANY EXPRESSED OR IMPLIED
* WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF
* MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED.
* IN NO EVENT SHALL COLLAB.NET OR ITS CONTRIBUTORS BE LIABLE FOR ANY
* DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
* DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE
* GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS
* INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER
* IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR
* OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF
* ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
*
* ====================================================================
*
* This software may consist of voluntary contributions made by many
* individuals on behalf of Collab.Net.
*/
#ifndef SVN_PARSE_H
#define SVN_PARSE_H
#include <svn_types.h>
#include <svn_string.h>
#include <svn_error.h>
#include <apr_pools.h>
#include <apr_hash.h>
#include <apr_file_io.h>
#include <ctype.h> /* isspace() */
/*
General Utility -- reads a config file, returns a hash of hashes.
For file format, see `notes/svn-config-files'
*/
svn_error_t * svn_parse (ap_hash_t **returnhash,
const char *filename,
ap_pool_t *pool);
/* Prints a hash, assuming all keys/vals are (svn_string_t *) */
void svn_hash_print (ap_hash_t *hash, FILE *stream);
/* Prints uberhash returned from svn_parse(),
assuming each key is (svn_string_t *)
and each val is printable by svn_hash_print(). */
void svn_uberhash_print (ap_hash_t *uberhash, FILE *stream);
#endif /* SVN_PARSE_H */
/*
* local variables:
* eval: (load-file "../svn-dev.el")
* end: */
------------- svn_parse.c
/* svn_parse: shared parsing routines for reading config files
*
* ================================================================
* Copyright (c) 2000 CollabNet. All rights reserved.
*
* Redistribution and use in source and binary forms, with or without
* modification, are permitted provided that the following conditions are
* met:
*
* 1. Redistributions of source code must retain the above copyright
* notice, this list of conditions and the following disclaimer.
*
* 2. Redistributions in binary form must reproduce the above copyright
* notice, this list of conditions and the following disclaimer in the
* documentation and/or other materials provided with the distribution.
*
* 3. The end-user documentation included with the redistribution, if
* any, must include the following acknowlegement: "This product includes
* software developed by CollabNet (http://www.Collab.Net)."
* Alternately, this acknowlegement may appear in the software itself, if
* and wherever such third-party acknowlegements normally appear.
*
* 4. The hosted project names must not be used to endorse or promote
* products derived from this software without prior written
* permission. For written permission, please contact info@collab.net.
*
* 5. Products derived from this software may not use the "Tigris" name
* nor may "Tigris" appear in their names without prior written
* permission of CollabNet.
*
* THIS SOFTWARE IS PROVIDED ``AS IS'' AND ANY EXPRESSED OR IMPLIED
* WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF
* MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED.
* IN NO EVENT SHALL COLLABNET OR ITS CONTRIBUTORS BE LIABLE FOR ANY
* DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
* DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE
* GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS
* INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER
* IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR
* OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF
* ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
*
* ====================================================================
*
* This software may consist of voluntary contributions made by many
* individuals on behalf of CollabNet.
*/
#include "svn_parse.h"
/*
NOT EXPORTED.
Input: an open file, a bytestring ptr, and a pool.
Returns: either APR_EOF or APR_SUCCESS, and a filled-in bytestring
containing the next line of the file.
(Note that the same bytestring can be reused in multiple
calls to this routine, because the bytestring is cleared at
the beginning.)
*/
apr_status_t
svn__my_readline (apr_file_t *file, svn_string_t *line, apr_pool_t *pool)
{
char c;
apr_status_t result;
svn_string_setempty (line); /* clear the bytestring first! */
while (1)
{
result = apr_getc (&c, file); /* read a byte from the file */
if (result == APR_EOF) /* file is finished. */
{
return APR_EOF;
}
if (c == '\n') /* line is finished. */
{
/* store the newline in our bytestring (important!) */
svn_string_appendbytes (line, &c, 1);
return APR_SUCCESS;
}
else /* otherwise, just append this byte to the bytestring */
{
svn_string_appendbytes (line, &c, 1);
}
}
}
/*
NOT EXPORTED.
Input: a bytestring, a handle for returning a bytestring, the
starting search offset, search character, and pool
Returns: 1. the offset of the search character (-1 if no match)
2. a newly allocated substring in "substr" (NULL if no match)
* This substring starts at `start' and goes to the offset
* This substring has no whitespace at either end
If used repeatedly, this routine is like a poor man's `split'
(combined with chomp).
*/
int
svn__slurp_to (const svn_string_t *searchstr,
svn_string_t **substr,
const size_t start,
const char sc,
apr_pool_t *pool)
{
int i;
/* Create a new bytestring */
*substr = svn_string_create ("<nobody home>", pool);
svn_string_setempty (*substr);
for (i = start; i < searchstr->len; i++)
{
if (searchstr->data[i] == sc)
{
/* printf ("found character '%c' at offset %d\n", sc, i);*/
svn_string_appendbytes (*substr, /* new substring */
searchstr->data + start,/* start copy */
(i - start)); /* number to copy */
svn_string_strip_whitespace (*substr);
return i;
}
}
/* If we get here, then the bytestring doesn't contain our search
character. This is bogus. */
*substr = NULL;
return -1;
}
/*
svn_parse() (finally)
Input: a filename and pool, pointer to a hash
Output: a pointer to a hash of hashes, all built within the pool
This routine parses a file which conforms to the standard
Subversion config file format (look in notes/).
The hash returned is a mapping from section-names to hash pointers;
each hash contains the keys/vals for each section. All
section-names, keys and vals are stored as svn_string_t pointers.
(These bytestrings are allocated in the same pool as the hashes.)
This routine makes no attempt to understand the sections, keys or
values. :) */
svn_error_t *
svn_parse (apr_hash_t **uberhash, const char *filename, apr_pool_t *pool)
{
apr_hash_t *current_hash; /* the hash we're currently storing vals in */
apr_pool_t *scratchpool;
svn_string_t *currentline;
apr_status_t result;
apr_file_t *file = NULL;
/* Create our uberhash */
*uberhash = apr_make_hash (pool);
/* Open the config file */
result = apr_open (&file,
filename,
APR_READ,
APR_OS_DEFAULT, /*TODO: WHAT IS THIS? */
pool);
if (result != APR_SUCCESS)
{
char *finalmsg; /* TODO: use apr's sprintf() thingie here */
svn_string_t *msg = svn_string_create
("svn_parse(): can't open for reading, file ", pool);
svn_string_appendstr (msg, filename);
return svn_error_create (result, NULL, NULL, pool, msg->data);
}
/* Create a scratch memory pool for buffering our file as we read it */
scratchpool = svn_pool_create (pool);
/* Create a bytestring to hold the current line of file */
currentline = svn_string_create ("<nobody home>", scratchpool);
/* Now start scanning our file, one line at a time */
while (svn__my_readline (file, currentline, scratchpool) != APR_EOF)
{
char c;
size_t offset = svn_string_first_non_whitespace (currentline);
if (offset == currentline->len)
{
/* whole line is whitespace, read next line! */
continue;
}
c = currentline->data[offset]; /* our first non-white character */
switch (c)
{
case '#':
{
/* It's a comment line, so read next line! */
continue;
};
case '[':
{
/* It's a new section! */
/* Create new hash to hold this section's keys/vals */
apr_hash_t *new_section_hash = apr_make_hash (pool);
/* Slurp up the section name */
svn_string_t *new_section;
svn__slurp_to (currentline, /* search current line */
&new_section, /* place new substring here */
offset + 1, /* start searching past the '[' */
']', /* look for this ending character */
pool); /* build our substring in this pool */
if (new_section == NULL) /* couldn't find a ']' ! */
{
policy->warning
(policy->data,
"svn_parse(): skipping malformed line: `%s'",
currentline->data, pool);
break;
}
/* printf ("Found new section: `");
svn_string_print (new_section, stdout, FALSE, FALSE);
printf ("'\n"); */
/* make this new hash the "active" hash for new keys/vals */
current_hash = new_section_hash;
/* store this new hash in our uberhash */
apr_hash_set (*uberhash,
new_section->data, /* key: bytestring */
new_section->len, /* the length of the key */
new_section_hash); /* val: ptr to the new hash */
break;
}
default:
{
/* If it's not a blank line, comment line, or section line,
then it MUST be a key : val line! */
/* Slurp up the key by searching for a colon */
svn_string_t *new_key, *new_val;
size_t local_offset;
local_offset = svn__slurp_to (currentline, /* search line */
&new_key, /* put substring here */
offset, /* start at this offset */
':', /* look for a colon */
pool); /* build substr here */
if (new_key == NULL) /* didn't find a colon! */
{
policy->warning
(policy->data,
"svn_parse(): skipping malformed line: `%s'",
currentline->data);
break;
}
/* Now slurp up the value, starting just past the colon */
svn__slurp_to (currentline,
&new_val,
local_offset + 1,
'\n',
pool);
/* printf ("Key: `");
svn_string_print (new_key, stdout, FALSE, FALSE);
printf ("'\n");
printf ("Val: `");
svn_string_print (new_val, stdout, FALSE, FALSE);
printf ("'\n"); */
/* Should we check for a NULL result from svn__slurp_to?
What are the chances it's not going to find a newline? :)
*/
/* Store key and val in the currently active hash */
apr_hash_set (current_hash,
new_key->data, /* key: bytestring data */
new_key->len, /* length of key */
new_val); /* val: ptr to bytestring */
break;
} /* default: */
} /* switch (c) */
} /* while (readline) */
/* Close the file and free our scratchpool */
result = apr_close (file);
if (result != APR_SUCCESS)
{
policy->warning
(policy->data,
"svn_parse(): can't close file: `%s'", filename->data);
}
apr_destroy_pool (scratchpool);
return SVN_NO_ERROR;
}
/* Convenience Routine: pretty-print an apr_hash_t.
(ASSUMING that all keys and vals are of type (svn_string_t *) )
*/
void
svn_hash_print (apr_hash_t *hash, FILE *stream)
{
apr_hash_index_t *hash_index; /* this represents a hash entry */
void *key, *val;
size_t keylen;
svn_string_t keystring, *valstring;
fprintf (stream, "\n-----> Printing hash:\n");
for (hash_index = apr_hash_first (hash); /* get first hash entry */
hash_index; /* NULL if out of entries */
hash_index = apr_hash_next (hash_index)) /* get next hash entry */
{
/* Retrieve key and val from current hash entry */
apr_hash_this (hash_index, &key, &keylen, &val);
/* Cast things nicely */
keystring.data = key;
keystring.len = keylen;
keystring.blocksize = keylen;
valstring = val;
/* Print them out nicely */
fprintf (stream, "Key: `");
/* TODO: svn_string_print (&keystring, stream, FALSE, FALSE); */
fprintf (stream, "', ");
fprintf (stream, "Val: `");
/* TODO: svn_string_print (valstring, stream, FALSE, FALSE); */
fprintf (stream, "'\n");
}
fprintf (stream, "\n");
}
/* Convenience Routine: pretty-print "uberhash" from svn_parse().
(ASSUMING that all keys are (svn_string_t *),
all vals are (apr_hash_t *) printable by svn_hash_print() )
*/
void
svn_uberhash_print (apr_hash_t *uberhash, FILE *stream)
{
apr_hash_index_t *hash_index; /* this represents a hash entry */
void *key, *val;
size_t keylen;
svn_string_t keystring;
apr_hash_t *valhash;
fprintf (stream, "\n-> Printing Uberhash:\n");
for (hash_index = apr_hash_first (uberhash); /* get first hash entry */
hash_index; /* NULL if out of entries */
hash_index = apr_hash_next (hash_index)) /* get next hash entry */
{
/* Retrieve key and val from current hash entry */
apr_hash_this (hash_index, &key, &keylen, &val);
/* Cast things nicely */
keystring.data = key;
keystring.len = keylen;
keystring.blocksize = keylen;
valhash = val;
/* Print them out nicely */
fprintf (stream, "---> Hashname: `");
/* TODO: svn_string_print (&keystring, stream, FALSE, FALSE); */
fprintf (stream, "'\n");
svn_hash_print (valhash, stream);
}
fprintf (stream, "\nUberhash printing complete.\n\n");
}
/*
* local variables:
* eval: (load-file "../svn-dev.el")
* end: */
------------- parsetest.c
/* Testing svn_parse() */
#include <stdio.h>
#include "svn_parse.h"
int
main ()
{
apr_hash_t *configdata;
apr_pool_t *pool;
svn_error_t *error;
/* Initialize APR (Apache pools) */
if (apr_initialize () != APR_SUCCESS)
{
printf ("apr_initialize() failed.\n");
exit (1);
}
pool = svn_pool_create (NULL);
/* Parse the file "./configfile" */
error =
svn_parse (&configdata, svn_string_create ("configfile", pool), pool);
if (error) {
svn_handle_error (error);
}
/* Print out our configdata uber-hash */
svn_uberhash_print (configdata, stdout);
/* If we were an application using libsvn_svr, we would now pass
this uber-hash to svn_init() to get a `svn_policies_t' structure.
We would then use this structure for all our wrappered filesystem
calls. */
/* Clean up our memory pool and apr */
apr_destroy_pool (pool);
apr_terminate ();
printf ("Test complete, exiting cleanly.\n\n");
return 0;
}
------------- configfile
# This is a sample Subversion config file
# Nicely Formatted Part
[Grades]
Joe : A
Mary : B-
Alice : A+
[Doughnuts]
Powdered : $.39
Chocolate : $1.59 a dozen
# BaDlY FoRMatTed paRT
[foosection]
giraffe : 17
elephant: green
# Yummy section
[unfinished section
[barsection ]
3.14159 :pi
eee : 2.71
ghoti : phish
# Really important section!
syntaxerror
[ bazsection ]
airplane: bahamas
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org
Re: String type madness
Posted by Branko Čibej <br...@xbc.nu>.
Branko �ibej wrote:
> Maybe. :-) I'll have a look at that code, and might use/expand it as
> necessary.
No I won't. This is from tigris:
Error: Unexpected output from cvs co:
*Check whether the directory
/usr/local/tigris/data/helm/cvs/repository/CVSROOT exists and the script
has write-access to the CVSROOT/history file if it exists.
The script needs to place lock files in the directory the file is in as
well.*
Ben, if you can get your hands on those two files, could you send them
to me, please?
Brane
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org
Re: String type madness
Posted by Tripp Lilley <tl...@perspex.com>.
On Thu, 26 Jul 2001, Branko [ISO-8859-2] �ibej wrote:
> Ben Collins-Sussman wrote:
>
> >Branko =?ISO-8859-2?Q?=C8ibej?= <br...@xbc.nu> writes:
> >
> >>Instead, I decided to resurrect a config file parsing library
...
> >I wrote a config-file parser just like this last summer for
> >Subversion, but then tossed it.
...
> Blimey. What's the world coming to, when everybody is stealing everyone
> else's ideas? :-)
Hey, I only wish I had running code, instead of just sketchy
architecture:
http://roml.sourceforge.net/
(actually, I'm being too hard on myself. I -do- have running code,
it's just too trivial to be released yet).
Anyway, I've just posted there a list of links to other
config-related tools I've found. Hopefully some of us busily
stealing each other's ideas :)
--
Tripp Lilley * http://stargate.eheart.sg505.net/~tlilley/
--------------------------------------------------------------------------
"The human soul is the greatest anti-piracy measure in the known
world, but no major company will use it."
<http://slashdot.org/comments.pl?sid=01/07/19/007240&cid=94>
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org
Re: String type madness
Posted by Branko Čibej <br...@xbc.nu>.
Ben Collins-Sussman wrote:
>Branko =?ISO-8859-2?Q?=C8ibej?= <br...@xbc.nu> writes:
>
>>Instead, I decided to resurrect a config file parsing library -- which /
>>should/ be in M3, IMHO -- I have lying around (It's LGPL, but I'm the
>>author, got a loopohole :-). I'm banging it into shape for Subversion,
>>
>
>
>FYI -- look in the Attic of libsvn_subr:
>
> http://subversion.tigris.org/source/browse/subversion/subversion/libsvn_subr/Attic/
>
>I wrote a config-file parser just like this last summer for
>Subversion, but then tossed it. It's already in our tree. And it
>parses config files exactly like our build.conf, exactly like win32
>.ini files. And there's even a test for it in libsvn_subr/parsetest/.
>It just returns a hash of hashes; the first hash key is the name of
>the [section], and it's value is another hash that contains all the
>key:val pairs.
>
Blimey. What's the world coming to, when everybody is stealing everyone
else's ideas? :-)
>I'm only pointing this out because adding a whole library for this
>might be heavy-handed. It might be easier just to move svn_parse.[ch]
>back into libsvn_subr. But then maybe your library does much more. :-)
>
Maybe. :-) I'll have a look at that code, and might use/expand it as
necessary.
Thanks for the pointer!
Brane
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org
Re: String type madness
Posted by Ben Collins-Sussman <su...@collab.net>.
Branko =?ISO-8859-2?Q?=C8ibej?= <br...@xbc.nu> writes:
> Instead, I decided to resurrect a config file parsing library -- which /
> should/ be in M3, IMHO -- I have lying around (It's LGPL, but I'm the
> author, got a loopohole :-). I'm banging it into shape for Subversion,
FYI -- look in the Attic of libsvn_subr:
http://subversion.tigris.org/source/browse/subversion/subversion/libsvn_subr/Attic/
I wrote a config-file parser just like this last summer for
Subversion, but then tossed it. It's already in our tree. And it
parses config files exactly like our build.conf, exactly like win32
.ini files. And there's even a test for it in libsvn_subr/parsetest/.
It just returns a hash of hashes; the first hash key is the name of
the [section], and it's value is another hash that contains all the
key:val pairs.
I'm only pointing this out because adding a whole library for this
might be heavy-handed. It might be easier just to move svn_parse.[ch]
back into libsvn_subr. But then maybe your library does much more. :-)
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org