[TYPO3-dev] redirect from domain vs. subdomain: different treatment - how to solve?

[grac]

PROBLEM SOLVED!


Franz Koch wrote:
> have you set the 'is root of website' flag for the page where the domain 
> record of the subdomain resides?
there had to be done a lot of screwing and twisting of the existing 
configuration, even replacing code of php files.

here my make-it-work - not being a crack it was more than just a pain 
and took me at least 3-5 days:


_"crawler_cli.phpsh":_

replace

line 29: define('PATH_thisScript',$_ENV['_']?$_ENV['_']:$_SERVER['_']);

by

line 29: $curf = 
'http://..................../typo3conf/ext/crawler/cli/crawler_cli.phpsh';
line 30: define('PATH_thisScript',$curf);

(thanks to diverse forum links)


_TSConfig:_

change

tx_crawler.crawlerCfg.paramSets {
    language = &L=[|_TABLE:pages_language_overlay;_FIELD:sys_language_uid]
    cHash = 1
    language.procInstrFilter = tx_indexedsearch_reindex, 
tx_indexedsearch_crawler
    baseUrl = http://............../
}
(wrong in http://wiki.typo3.org/Crawler!)

tx_crawler.crawlerCfg.paramSets {
  tt_news = &tx_ttnews[tt_news]=[_TABLE:tt_news;_PID:xxx]
  tt_news.procInstrFilter = tx_indexedsearch_reindex, tx_cachemgm_recache
  tt_news.cHash = 1
  tt_news.pidsOnly = yyy
  baseUrl = http://............../
}

to

tx_crawler.crawlerCfg.paramSets {
    language = &L=[|_TABLE:pages_language_overlay;_FIELD:sys_language_uid]
    cHash = 1
    language.procInstrFilter = tx_indexedsearch_reindex, 
tx_indexedsearch_crawler
    language.baseUrl = http://............../
}

tx_crawler.crawlerCfg.paramSets {
  tt_news = &tx_ttnews[tt_news]=[_TABLE:tt_news;_PID:xxx]
  tt_news.procInstrFilter = tx_indexedsearch_reindex, tx_cachemgm_recache
  tt_news.cHash = 1
  tt_news.pidsOnly = yyy
  tt_news.baseUrl = http://............../
}

i.e. in every config-set the respective [key] has to be set before baseUrl.
(this is the only guessable point in the manual of "crawler" for this 
make-it-work )

exeption (not in the manual):

tx_crawler.crawlerCfg.paramSets.grp1 < tx_crawler.crawlerCfg.paramSets
tx_crawler.crawlerCfg.paramSets.grp1{
    userGroups = 0,-2,x
}

here is baseUrl not needed at all.



_TStemplate / setup:_

config.redirectOldLinksToNew = 0

(for cooluri, thanks to http://wiki.typo3.org/Crawler)



with this configuration the crawler runs directly more or less without 
errors and the cronscript doesn't send CRON DAEMON messages any more.

this also means that my original guess that there might be a difference 
between redirected domains and redirected subdomains was wrong.

but i'm very angry at and shocked about the author's arrogance not 
having written these necessary steps in his manual (EXT version 2.1.0, 
2007).
especially the necessary modification of the initialising php file .
which normal gifted admin would think of a necessary deeper look into 
this file when you find in the header the following statement:

line 12: ## DEPRICATED - instead, use typo3/cli_dispatch.php, follow 
instructions in there.

typo3 initially meant for me that it's enough work to familiarise with 
TS, TSConfig and everything around.
going deep into php and changing php files - i thought - was a matter 
for developers.
which was obviously a mistake.


please update at least the manual and make the information for the 
necessary change in "crawler_cli.phpsh" accessible for admins!

thanks
stephan