mirror of
https://github.com/yacy/yacy_search_server.git
synced 2024-09-19 00:01:41 +02:00
- all parsers are activated by default for pro releases
- slightly higher file size limits for parsers git-svn-id: https://svn.berlios.de/svnroot/repos/yacy/trunk@4051 6c8d7289-2bf4-0310-a012-ef5d649a1542
This commit is contained in:
parent
694defb257
commit
0ad8499e66
|
@ -602,34 +602,7 @@ public class SettingsAck_p {
|
|||
prop.put("info_parser",enabledMimesCount);
|
||||
prop.put("info", 18);
|
||||
return prop;
|
||||
|
||||
// plasmaSwitchboard sb = (plasmaSwitchboard)env;
|
||||
//
|
||||
// HashMap configList = plasmaParser.getParserConfigList();
|
||||
// Iterator parserModeIter = configList.keySet().iterator();
|
||||
//
|
||||
// String[] enabledMimes = null;
|
||||
// if (post.containsKey("allParserEnabled")) {
|
||||
// // enable all available parsers
|
||||
// enabledMimes = plasmaParser.setEnabledParserList(sb.parser.getAvailableParserList().keySet());
|
||||
// } else {
|
||||
// // activate all received parsers
|
||||
// enabledMimes = plasmaParser.setEnabledParserList(post.keySet());
|
||||
// }
|
||||
// Arrays.sort(enabledMimes);
|
||||
//
|
||||
// StringBuffer enabledMimesTxt = new StringBuffer();
|
||||
// for (int i=0; i < enabledMimes.length; i++) {
|
||||
// enabledMimesTxt.append(enabledMimes[i]).append(",");
|
||||
// prop.put("info_parser_" + i + "_enabledMime",enabledMimes[i]);
|
||||
// }
|
||||
// prop.put("info_parser",enabledMimes.length);
|
||||
// if (enabledMimesTxt.length() > 0) enabledMimesTxt.deleteCharAt(enabledMimesTxt.length()-1);
|
||||
//
|
||||
// env.setConfig("parseableMimeTypes",enabledMimesTxt.toString());
|
||||
//
|
||||
// prop.put("info", 18);
|
||||
// return prop;
|
||||
|
||||
}
|
||||
|
||||
// Crawler settings
|
||||
|
|
19
yacy.init
19
yacy.init
|
@ -186,7 +186,7 @@ proxyCache = DATA/HTCACHE
|
|||
|
||||
# the maximum disc cache size for files in proxyCache in megabytes
|
||||
proxyCacheSize = 100
|
||||
proxyCacheSize__pro = 600
|
||||
proxyCacheSize__pro = 1024
|
||||
|
||||
# storage place for new releases
|
||||
releases = DATA/RELEASE
|
||||
|
@ -216,10 +216,15 @@ proxyCacheMigration = true
|
|||
# parseableMime: specifies mime-types that can be indexed but not on the fly
|
||||
parseableRealtimeMimeTypes=application/xhtml+xml,text/html,text/plain,text/sgml
|
||||
parseableMimeTypes=
|
||||
parseableMimeTypes__pro=application/atom+xml,application/bzip2,application/excel,application/gzip,application/java-archive,application/msexcel,application/mspowerpoint,application/msword,application/octet-stream,application/pdf,application/postscript,application/powerpoint,application/rdf+xml,application/rss+xml,application/rtf,application/tar,application/vcard,application/vnd.ms-excel,application/vnd.ms-powerpoint,application/vnd.oasis.opendocument.text,application/x-7z-compressed,application/x-bz2,application/x-bzip2,application/x-compress,application/x-compressed,application/x-excel,application/x-gzip,application/x-msexcel,application/x-redhat packet manager,application/x-redhat-package-manager,application/x-rpm,application/x-shockwave-flash,application/x-shockwave-flash2-preview,application/x-tar,application/x-vnd.oasis.opendocument.text,application/x-xml,application/x-zip,application/x-zip-compressed,application/xml,application/zip,text/postscript,text/rss,text/rtf,text/x-vcard,text/xml
|
||||
parseableMimeTypes.CRAWLER=
|
||||
parseableMimeTypes.CRAWLER__pro=application/atom+xml,application/bzip2,application/excel,application/gzip,application/java-archive,application/msexcel,application/mspowerpoint,application/msword,application/octet-stream,application/pdf,application/postscript,application/powerpoint,application/rdf+xml,application/rss+xml,application/rtf,application/tar,application/vcard,application/vnd.ms-excel,application/vnd.ms-powerpoint,application/vnd.oasis.opendocument.text,application/x-7z-compressed,application/x-bz2,application/x-bzip2,application/x-compress,application/x-compressed,application/x-excel,application/x-gzip,application/x-msexcel,application/x-redhat packet manager,application/x-redhat-package-manager,application/x-rpm,application/x-shockwave-flash,application/x-shockwave-flash2-preview,application/x-tar,application/x-vnd.oasis.opendocument.text,application/x-xml,application/x-zip,application/x-zip-compressed,application/xml,application/zip,text/postscript,text/rss,text/rtf,text/x-vcard,text/xml
|
||||
parseableMimeTypes.PROXY=
|
||||
parseableMimeTypes.PROXY__pro=application/atom+xml,application/bzip2,application/excel,application/gzip,application/java-archive,application/msexcel,application/mspowerpoint,application/msword,application/octet-stream,application/pdf,application/postscript,application/powerpoint,application/rdf+xml,application/rss+xml,application/rtf,application/tar,application/vcard,application/vnd.ms-excel,application/vnd.ms-powerpoint,application/vnd.oasis.opendocument.text,application/x-7z-compressed,application/x-bz2,application/x-bzip2,application/x-compress,application/x-compressed,application/x-excel,application/x-gzip,application/x-msexcel,application/x-redhat packet manager,application/x-redhat-package-manager,application/x-rpm,application/x-shockwave-flash,application/x-shockwave-flash2-preview,application/x-tar,application/x-vnd.oasis.opendocument.text,application/x-xml,application/x-zip,application/x-zip-compressed,application/xml,application/zip,text/postscript,text/rss,text/rtf,text/x-vcard,text/xml
|
||||
parseableMimeTypes.ICAP=
|
||||
parseableMimeTypes.ICAP__pro=application/atom+xml,application/bzip2,application/excel,application/gzip,application/java-archive,application/msexcel,application/mspowerpoint,application/msword,application/octet-stream,application/pdf,application/postscript,application/powerpoint,application/rdf+xml,application/rss+xml,application/rtf,application/tar,application/vcard,application/vnd.ms-excel,application/vnd.ms-powerpoint,application/vnd.oasis.opendocument.text,application/x-7z-compressed,application/x-bz2,application/x-bzip2,application/x-compress,application/x-compressed,application/x-excel,application/x-gzip,application/x-msexcel,application/x-redhat packet manager,application/x-redhat-package-manager,application/x-rpm,application/x-shockwave-flash,application/x-shockwave-flash2-preview,application/x-tar,application/x-vnd.oasis.opendocument.text,application/x-xml,application/x-zip,application/x-zip-compressed,application/xml,application/zip,text/postscript,text/rss,text/rtf,text/x-vcard,text/xml
|
||||
parseableMimeTypes.URLREDIRECTOR=
|
||||
parseableMimeTypes.URLREDIRECTOR__pro=application/atom+xml,application/bzip2,application/excel,application/gzip,application/java-archive,application/msexcel,application/mspowerpoint,application/msword,application/octet-stream,application/pdf,application/postscript,application/powerpoint,application/rdf+xml,application/rss+xml,application/rtf,application/tar,application/vcard,application/vnd.ms-excel,application/vnd.ms-powerpoint,application/vnd.oasis.opendocument.text,application/x-7z-compressed,application/x-bz2,application/x-bzip2,application/x-compress,application/x-compressed,application/x-excel,application/x-gzip,application/x-msexcel,application/x-redhat packet manager,application/x-redhat-package-manager,application/x-rpm,application/x-shockwave-flash,application/x-shockwave-flash2-preview,application/x-tar,application/x-vnd.oasis.opendocument.text,application/x-xml,application/x-zip,application/x-zip-compressed,application/xml,application/zip,text/postscript,text/rss,text/rtf,text/x-vcard,text/xml
|
||||
|
||||
# media extension string
|
||||
# a comma-separated list of extensions that denote media file formats
|
||||
|
@ -697,16 +702,16 @@ onlineCautionDelay=10000
|
|||
# Some configuration values for the crawler
|
||||
crawler.clientTimeout=9000
|
||||
|
||||
# http crawler specific settings
|
||||
# http crawler specific settings; size in bytes
|
||||
crawler.http.acceptEncoding=gzip
|
||||
crawler.http.acceptLanguage=en-us,en;q=0.5
|
||||
crawler.http.acceptCharset=ISO-8859-1,utf-8;q=0.7,*;q=0.7
|
||||
crawler.http.maxFileSize=131072
|
||||
crawler.http.maxFileSize__pro=524288
|
||||
crawler.http.maxFileSize=262144
|
||||
crawler.http.maxFileSize__pro=1048576
|
||||
|
||||
# ftp crawler specific settings
|
||||
crawler.ftp.maxFileSize=131072
|
||||
crawler.ftp.maxFileSize__pro=524288
|
||||
# ftp crawler specific settings; size in bytes
|
||||
crawler.ftp.maxFileSize=262144
|
||||
crawler.ftp.maxFileSize__pro=1048576
|
||||
|
||||
# maximum number of crawler threads
|
||||
crawler.MaxActiveThreads = 30
|
||||
|
|
Loading…
Reference in New Issue
Block a user