Обсуждение: Texinfo docs/target
I've been using a texinfo (more precisely: info) version of the docs for some weeks now, and really came to love it. It is still the version I mentioned here: <http://archives.postgresql.org/pgsql-general/2005-10/msg01813.php> Now I'm wondering, if it'd be worth fixing the index and some minor encoding deficiencies (e.g. medium-length dashes end up as "[mdash]"), and adding Autoconf support in order to make it a proper target. I remember other requests for info-docs on the lists than the above, but it is rather hard to search for them :-/. regards, Andreas --
Andreas Seltenreich wrote: > Now I'm wondering, if it'd be worth fixing the index and some minor > encoding deficiencies (e.g. medium-length dashes end up as > "[mdash]"), and adding Autoconf support in order to make it a proper > target. We probably won't ship it with the default distribution, but I don't see any harm in at least providing build support for it initially. -- Peter Eisentraut http://developer.postgresql.org/~petere/
Peter Eisentraut <peter_e@gmx.net> writes:
> Andreas Seltenreich wrote:
>> Now I'm wondering, if it'd be worth fixing the index and some minor
>> encoding deficiencies (e.g. medium-length dashes end up as
>> "[mdash]"), and adding Autoconf support in order to make it a proper
>> target.
>
> We probably won't ship it with the default distribution, but I don't see
> any harm in at least providing build support for it initially.
The index seems to work with more recent versions of DB2X, and the
apparent encoding problems turned out to be an overlook in the
postgres.xml target.
I had to run makeinfo with the --no-validate option, mostly because
the info design doesn't allow duplicate node names. However, it
doesn't seem to impair navigation besides addressing nodes by name,
and since most of the duplicate names are nodes like "Changes" in the
release notes section, I don't think it will be more of a nuisance
than being unable to look up Pi in the index...
I figured adding autoconf support for db2x/info wouldn't be very wise
as long as the experimental postgres.xml target the new rules depend
on doesn't have any either.
Patch attached.
regards,
andreas
Index: doc/src/sgml/Makefile
===================================================================
RCS file: /projects/cvsroot/pgsql/doc/src/sgml/Makefile,v
retrieving revision 1.76
diff -u -r1.76 Makefile
--- doc/src/sgml/Makefile 6 Apr 2006 18:54:36 -0000 1.76
+++ doc/src/sgml/Makefile 29 Sep 2006 04:27:59 -0000
@@ -212,7 +212,7 @@
postgres.xml: postgres.sgml $(GENERATED_SGML)
$(OSX) -x lower $< | \
- sed -e 's/\[\(lt\|gt\|amp\|nbsp\|copy\|quot\|ouml\|uuml\|egrave\) *\]/\&\1;/g' \
+ sed -e 's/\[\(amp\|copy\|egrave\|gt\|lt\|mdash\|nbsp\|ouml\|pi\|quot\|uuml\) *\]/\&\1;/g' \
-e '1a\' -e '<!DOCTYPE book PUBLIC "-//OASIS//DTD DocBook XML V4.2//EN"
"http://www.oasis-open.org/docbook/xml/4.2/docbookx.dtd">'\
>$@
@@ -222,6 +222,24 @@
htmlhelp: stylesheet-hh.xsl postgres.xml
$(XSLTPROC) $(XSLTPROCFLAGS) --stringparam pg.version '$(VERSION)' $^
+##
+## Experimental Texinfo targets
+##
+
+DB2X_TEXIXML = db2x_texixml
+DB2X_XSLTPROC = db2x_xsltproc
+MAKEINFO = makeinfo
+
+postgres.texixml: postgres.xml
+ $(DB2X_XSLTPROC) -s texi $< -o $@
+
+postgres.texi: postgres.texixml
+ $(DB2X_TEXIXML) --encoding=iso-8859-1//TRANSLIT $< \
+ --to-stdout > $@
+
+postgres.info: postgres.texi
+ $(MAKEINFO) --enable-encoding --no-split \
+ --no-validate $< -o $@
##
## Check
@@ -249,3 +267,5 @@
rm -f INSTALL HISTORY regress_README
# XML
rm -f postgres.xml htmlhelp.hhp toc.hhc index.hhk
+# Texinfo
+ rm -f *.texixml *.texi *.info db2texi.refs
Your patch has been added to the PostgreSQL unapplied patches list at:
http://momjian.postgresql.org/cgi-bin/pgpatches
It will be applied as soon as one of the PostgreSQL committers reviews
and approves it.
---------------------------------------------------------------------------
Andreas Seltenreich wrote:
> Peter Eisentraut <peter_e@gmx.net> writes:
>
> > Andreas Seltenreich wrote:
> >> Now I'm wondering, if it'd be worth fixing the index and some minor
> >> encoding deficiencies (e.g. medium-length dashes end up as
> >> "[mdash]"), and adding Autoconf support in order to make it a proper
> >> target.
> >
> > We probably won't ship it with the default distribution, but I don't see
> > any harm in at least providing build support for it initially.
>
> The index seems to work with more recent versions of DB2X, and the
> apparent encoding problems turned out to be an overlook in the
> postgres.xml target.
>
> I had to run makeinfo with the --no-validate option, mostly because
> the info design doesn't allow duplicate node names. However, it
> doesn't seem to impair navigation besides addressing nodes by name,
> and since most of the duplicate names are nodes like "Changes" in the
> release notes section, I don't think it will be more of a nuisance
> than being unable to look up Pi in the index...
>
> I figured adding autoconf support for db2x/info wouldn't be very wise
> as long as the experimental postgres.xml target the new rules depend
> on doesn't have any either.
>
> Patch attached.
>
> regards,
> andreas
>
>
> ---------------------------(end of broadcast)---------------------------
> TIP 4: Have you searched our list archives?
>
> http://archives.postgresql.org
--
Bruce Momjian bruce@momjian.us
EnterpriseDB http://www.enterprisedb.com
+ If your life is a hard drive, Christ can be your backup. +
Patch applied. Thanks. --------------------------------------------------------------------------- Andreas Seltenreich wrote: > Peter Eisentraut <peter_e@gmx.net> writes: > > > Andreas Seltenreich wrote: > >> Now I'm wondering, if it'd be worth fixing the index and some minor > >> encoding deficiencies (e.g. medium-length dashes end up as > >> "[mdash]"), and adding Autoconf support in order to make it a proper > >> target. > > > > We probably won't ship it with the default distribution, but I don't see > > any harm in at least providing build support for it initially. > > The index seems to work with more recent versions of DB2X, and the > apparent encoding problems turned out to be an overlook in the > postgres.xml target. > > I had to run makeinfo with the --no-validate option, mostly because > the info design doesn't allow duplicate node names. However, it > doesn't seem to impair navigation besides addressing nodes by name, > and since most of the duplicate names are nodes like "Changes" in the > release notes section, I don't think it will be more of a nuisance > than being unable to look up Pi in the index... > > I figured adding autoconf support for db2x/info wouldn't be very wise > as long as the experimental postgres.xml target the new rules depend > on doesn't have any either. > > Patch attached. > > regards, > andreas > > > ---------------------------(end of broadcast)--------------------------- > TIP 4: Have you searched our list archives? > > http://archives.postgresql.org -- Bruce Momjian bruce@momjian.us EnterpriseDB http://www.enterprisedb.com + If your life is a hard drive, Christ can be your backup. +
Bruce Momjian writes:
> Patch applied. Thanks.
Here's a small update to pass the basename of the target as a
stylesheet parameter. This makes the resulting info file work
properly when compressed or split. It also prevents Debian's
"install-info" script from choking.
Thanks,
andreas
Index: doc/src/sgml/Makefile
===================================================================
RCS file: /projects/cvsroot/pgsql/doc/src/sgml/Makefile,v
retrieving revision 1.77
diff -u -r1.77 Makefile
--- doc/src/sgml/Makefile 3 Oct 2006 21:13:17 -0000 1.77
+++ doc/src/sgml/Makefile 5 Oct 2006 00:32:33 -0000
@@ -231,7 +231,7 @@
MAKEINFO = makeinfo
postgres.texixml: postgres.xml
- $(DB2X_XSLTPROC) -s texi $< -o $@
+ $(DB2X_XSLTPROC) -s texi -g output-file=$(basename $@) $< -o $@
postgres.texi: postgres.texixml
$(DB2X_TEXIXML) --encoding=iso-8859-1//TRANSLIT $< \
Andreas Seltenreich wrote: > Bruce Momjian writes: > > > Patch applied. Thanks. > > Here's a small update to pass the basename of the target as a > stylesheet parameter. This makes the resulting info file work > properly when compressed or split. It also prevents Debian's > "install-info" script from choking. Applied. Thanks. -- Bruce Momjian bruce@momjian.us EnterpriseDB http://www.enterprisedb.com + If your life is a hard drive, Christ can be your backup. +