# /etc/gorg/gorg.conf: Configuration file for Gorg

# Root dir, typically, your DocumentRoot 
# (f)cgi scripts find it in their environment but
# the stand-alone webserver and the search engine need it
root = "/home/neysx/gentoo.org/gentoo/xml/htdocs"

# Make webrick listen on given IP (IP onlyu, no host name)
listen = 127.0.0.1

# Mount paths that are not under the root directory (used by stand-alone web server only)
# eg. to mount /cgi-bin which is usually not under the document root
# Note: Those directories will be handled by the stock FileHandler, ie. not by gorg
mount = /cgi-bin on /home/neysx/gentoo.org/gentoo/xml/cgi-bin
mount = /images on /home/neysx/gentoo.org/gentoo/xml/images

# Should gorg accept cookies and pass $param=$value to the xsl transform
# Default is no (anything but 1 is no)
acceptCookies = 1

# Only read so many lines in xml files to identify stylesheets, use 0 to scan whole file 
headXSL = 12

# Default stylesheet, relative to root dir
defaultXSL = "/xsl/guide.xsl"

# Only used by fastCGI, auto exit after given number of requests (0 means no)
# The fcgi process manager will restart a new instance automatically
# NB: it will NOT exit before at least 1 full minute has elapsed even if you set a very low value
# If you want a really short-lived version, use the cgi instead
# mod_fcgid does its own process recycling and this feature will be obsoleted in an later version
autoKill = 5000

# Allow return of unprocessed xml file if passthru==(anything but 0) appears in URI params
# 0==No, anything else==Yes
passthru = 1

# Pass pathname of requested file in named parameter to the xsl transform
# [a-zA-Z]+ , anything else is ignored and no param is passed
# Default is "link"
linkParam = link

# Pass a param named httphost to the style sheet (== host value from HTTP header)
# 0 or nothing (default) disables this feature
# * will pass the value as received from the user agent or none (http/1.0)
# name alias1 alias2... will pass name when the value sent by the user agent
#                       matches exactly any of name alias1 alias2...
#                       if any alias is *, any value (even nil) will match and name will be passed
# When no value matches, the value received from the user agent is passed
#httphost = mysite www.mysite.org mysite.org alias.mysite.org

# Cache directory. Directory must exist and be writable by whoever runs the server (e.g. apache)
# It must also be writable by the user who runs the stand-alone web server, i.e. not the apache user
# if you want to use both web servers. You can even run both at the same time.
# Default is no cache
cacheDir = "/var/cache/gorg"

# Number of seconds after which a document is considered too old, 0=never
# Can be used to force a refresh or to stress-test the system
#cacheTTL = 86400 # 1 day
cacheTTL = 864000 # or 10 days
#cacheTTL = 600 # or 10 minutes....

# Use a tree of directories under cacheDir that matches the site tree
# Use when your system has problems coping with a huge single cache dir
# 0 means no tree (all files in cacheDir) and is the default
# If you use this, make sure you clean up the cache with gorg -C regularly
cacheTree = 1

# Max size of cache in megabytes
# Please note that cacheSize is used ONLY when cleaning up either
#    when cacheTree==0 and a clean-up is started based on cacheWash (see below)
# or when cacheTree!=0 and `gorg -C` is run
cacheSize = 250

# Max number of files in a given cache directory
# Please note that this limit is also enforced when cacheTree == 0
# in which case it means the max total number of files in the whole cache
maxFiles = 2000

# Support gzip http encoding (ie. mod_deflate)
# 0 means no compression *and* no support for gzip encoding.
# 1-9 gives compression level, 1 least compressed, 9 max compressed
# Cached pages use the same compression level
# Default is 2
zipLevel = 2

# Clean cache automatically and regularly when a store into the cache occurs.
# gorg cleans up if random(value) < 10, i.e.
# Set to 0 to disable and rely on gorg --clean-cache being run regularly
# a value<=10 means at every call,
# 100 means 10 percent of stores will also clean the cache
# 1000 means 10 permille (cacheSize will be checked only once every 100 stores)
# Note: gorg only tries to clean the dir it caches to, not the whole cache tree
# Use `gorg -C` or `gorg --clean-cache` to clean up whole cache
cacheWash = 0

# Level of logging that goes to syslog
# OFF, FATAL, ERROR, WARN, INFO, DEBUG = 0, 1, 2, 3, 4, 5
logLevel = 4

#
# Used only by stand-alone webserver
#

# Send hit stats to syslog/stderr/a file
accessLog = "syslog"

# Listen on port (must be >1023 to be run by non-root)
port = 8008

#
# Search engine parameters
#

# Connect string, only mysql is supported at the moment
dbConnect  = DBI:mysql:DB_NAME:HOST_NAME
dbUser     = USENAME
dbPassword = PASSWORD

# Document language can be guessed from the document itself with
# an XPath expression. It should return the language code.
# Only the first 5 characters will be used.
# For instance, to use a root element's lang attribute:
xpath_to_lang = /*[1]/@lang

# If no XPath is given or no lang is found, you can use the file path as a fallback:
# define a regexp to apply to the file path, $1 must yield the language
# For instance, the following one applied to '/doc/en/file.xml' returns 'en'
fpath_to_lang = ^/[^/]+/([^/]+)/.*xml$

# include/exclude directives will be processed in the order they appear below.
# First match will be used to either include or exclude the file.
# If no match is found, file is skipped
# Each directive should define one and only one regexp
# Beware, regexp are not shell globs, .xml means any character followed by xml anywhere in the file name
# .+\.xml$  means one or more characters followed by a dot and ending with xml
# Any file that can't be processed, ie. because it is not well-formed will not be indexed

exclude = ^/proj/en/gdp/tests/
exclude = /CVS/
exclude = ^/xsl/
exclude = /draft/
exclude = ^/doc/.+/handbook/2004
exclude = metadoc\.xml$
exclude = /inserts-.+\.xml$
exclude = ^/dyn/
exclude = herds/pkgList.xml
include = ^/.+\.xml$