As reported edycop in mantis 765 (
http://mantis.tokeek.de/view.php?id=765 ), parsing of xlsx files was
quite incomplete.
Now properly support "Shared String Table" entry in Office Open XML
spreadsheets, an also detect embedded URLs.
Integrating the Apache poi-ooxml library could be an option for finer
OOXML formats support, but their SAX style parsing example (
http://poi.apache.org/spreadsheet/how-to.html#xssf_sax_api ) tends to
show that a custom SAX handler is still efficient for lightweight and
low memory footprint processing.
Some web servers provide both 'Content-Encoding : "gzip"' and
'Content-Type : "application/x-gzip"' HTTP headers on their ".gz" files.
This was annoying to fail on such resources which are not so uncommon,
while non conforming (see RFC 7231 section 3.1.2.2 for
"Content-Encoding" header specification
https://tools.ietf.org/html/rfc7231#section-3.1.2.2)
Thus enable getpageinfo_p API to return something in a reasonable amount
of time on resources over MegaBytes size range.
Support added first with the generic XML parser, for other formats
regular crawler limits apply as usual.
Recursive processing was removed in commit
67beef657f, but one remained for anchors
content(likely omitted from refactoring). It is no more necessary :
other links such as images embedded in anchors are currently correctly
detected by the parser.
More annoying : that remaining recursive processing could lead to almost
endless processing when encountering some (invalid) HTML structures
involving nested anchors, as detected and reported by lucipher on YaCy
forum ( http://forum.yacy-websuche.de/viewtopic.php?f=23&t=6005 ).
For faster processing (measured about 2 times faster on many real-world
examples) and more advanced detection (previous algorithm detected only
URLs separated from the rest of the text by a space character).
Especially for Turkish speaking users using "tr" as their system default
locale : strings for technical stuff (URLs, tag names, constants...)
must not be lower cased with the default locale, as 'I' doesn't becomes
'i' like in other locales such as "en", but becomes 'ı'.
This parser adds support for any XML based format other than already
supported XML vocabularies such XHTML, RSS/Atom feeds... It will
eventually be used as a fallback if one of these specific parsers fail,
before falling back to the existing genericParser which extracts not
that much useful information except URL tokens.
Also add when possible a warning level log message on input stream
closing error instead of failing silently. This could help understanding
some IO exceptions such as "too many files open".
Revealed by commit c77e43a : the exception was then thrown when indexing
pages containing mailto: scheme URL links with the Solr Webgraph core
enabled.
Fixed the error case and restored filtering on mailto links in
Document.resortLinks() as these URLs still should not appear in
Document.hyperlinks.
On MediaWiki dump imports, the SurrogateReader was trying to unread too
many bytes, then failing with the following exception :
"java.io.IOException: Push back buffer is full".
When using a public HTTP URL in /IndexImportMediawiki_p.html, the remote
file now is directly streamed and processed, allowing import of several
GB dumps even with a low memory remote peer, and without need to
manually download the dump file first.
warc = Web ARChive File Format.
Warc files with extension .warc or compressed warc.gz can be placed in the
DATA/surrogate/in and contained responses are imported to the index.
The used library is stream based so we can easily extend it later to use
and load warc's from the net.
by using icu.ULocale for languages not already covered (ICU normalizes
to ISO639-1 2 char codes).
Add test class
Use DublinCore vocabulary declarations in DCEntry and SurrogateReader
for easier usage debugging,
Init SurrogateReader.inputSource on first use.
recognized as tag like 1<a
reported in https://github.com/yacy/yacy_search_server/issues/109
Script content is ignored by default, but the text is filtered for html
tags. Modified scraper to skip tag filtering while within a <script>
section (until a closing tag is detected </script>.
Possible side effect, missing </script> end-tag will truncate trailing
content text.
add "datetime" property of <time> tag to scrapers startdate list.
Datetime is parsed as iso8601 (xml) date, html5 allows partial as well
as duration (not handled by this)
NullPointerException occurred when using and Identificator instance
which encountered and error in its constructor.
This error could be caused by a missing "langdetect" folder in the
current folder of the main process, or by simultaneous first calls to
the constructor, initializing concurrently the DetectorFactory.langlist.
Fixes the mantis 714 (http://mantis.tokeek.de/view.php?id=714)
Could occur when a search request was performed just after peer startup,
and the Switchboard Thread "LibraryProvider.initialize" had completed,
thus requesting a ProbabilisticClassifier not completely initialized
(and having a null contexts property).
Applied strategy : when there is no restriction on domains or
sub-path(s), stack anchor links once discovered by the content scraper
instead of waiting the complete parsing of the file.
This makes it possible to handle a crawling start file with thousands of
links in a reasonable amount of time.
Performance limitation : even if the crawl start faster with a large
file, the content of the parsed file still is fully loaded in memory.
to make sure current dates are recognized (was fixed to 2014 - 2016)
+ adjust holiday date parser from pattern.match to pattern.find to deal with leading and trailing text
+ moved relative date recognition (morgen, tomorrow) to parseline (used by query parser only), as not working and problematic for indexing
+ add test case for parseline (used by query parser)
including a small change, word posintext counting.
We remember/store 1st posintext. Previously following words got a handle (posintext)
excluding found. Now it just counts and assigns true posintext as handle (posintext)
to also support handling of urls w/o corresponding file-extension.
For this refactor use of document.getParserObject() to alway return a Parser (for clean logic)
and define/move the scraperObject as local var of AbstractParser.
Adjust related calls to getParserObject (where actually a scraperObject is wanted).
Addionally skip appending url token to parsed text for dht metadata entries
(by default returned as result by rwi index).
- Above brought up that parser start url parameter, declared as AnchorURL uses only methodes of parent object DigestURL (changed parameter declaration accordingly).
Language identification may show poor performance on documents with short or no
title but clear lang indication in text content. Using content text too
improves lang detection.
+ remove double caching of text in Identificator
in worker thread.
Writer of importer keeps needs a poison to close the file. On exception (e.g. OOM)
add a poison marker in outer most try/catch to assure output queue will terminate
in this condition too (and closes+renames the surrogate/in/xxx.prt file)
luccioman