Marcin,
I am nto sure about English, since there seem to be variants of British,
USA, Canadian, Zew Zealand, Australia, that could be considered
local/regional variants. A country is just a region of the world, right?
Or is this too philosophical?
Ruud
On 07-10-13 16:03, Marcin Miłkowski
No, those are the postag dictionaries!
Ruud
On 05-10-13 13:15, Stefan Lotties wrote:
LO/OO provides dictionaries when installing it. Can't we use them
instead of our own hunspell dictionaries? Languages such as german
stiff have a huge grammar.xml beside the dictionary, but leaving out
the
Right. It crashed again with the big batch.
Some solution would be nice.
Ruud
On 02-10-13 12:00, Stefan Lotties wrote:
On Wed, Oct 2, 2013 at 11:34 AM, Daniel Naber list2...@danielnaber.de wrote:
On 2013-10-02 11:14, Stefan Lotties wrote:
Sounds right. The fix should look very similar to
- with it
It could be platform, But I am not the only one on (K)ubuntu, using JDK.
Ruud
On 2013-09-30 18:17, Ruud Baars wrote:
For reproducing exactly: remove the line feeds from the sentence. They
were introduced by the e-mail.
That and using your disambiguation file didn't help, I
Pellé wrote:
Ruud Baars wrote:
I retrace the steps with 2.3, and the error does not reproduce. It must
be a snapshot thing.
No matter, case closed.
Ruud
Hopefully it's fixed indeed... or it's a rare multi-threading bug
and those are painful bugs to debug and reproduce. In fact
I just saw
I just added a bug for this on Sourceforge.
I was wondering what all those fields the creator is able to set are
doing there, there is no explanation at all.
What is Milestone?
Owner?
Which priority is highest?
Should all these items not be set by the 'moderator' of the bugs list?
Ruud
, echoing the file done into a admin file, so I can find out in
what file it crashes.
Ruud
On 29-09-13 12:32, Daniel Naber wrote:
On 2013-09-29 11:02, Ruud Baars wrote:
But I will have do redo it all (it happend after 14 GB of output ..) ,
and there is no indication of file it was in, nor
deletion or Undo addition.
Regards,
Panagiotis Minos
On 29/09/2013 01:04 ??, Ruud Baars wrote:
I guess this is the same thing I reported, and is caused by missing
translations of Java itself.
There apperently is a trick to work around it, but it is too late,
bevause of feature freeze ..
Ruud
Can or should I add an exception to this disambiguation rule (in fact a
postag assignment) to prevent a word to get the postag NN1r twice, or is
that not necessary?
Can I add exceptions to the disambig rule?
Ruud
rule name=onbekend verkleinwoord enkelvoud id=UNKNOWN_NN1r
pattern
Might be a good idea to add after the release ..
Ruud
On 25-09-13 08:45, Daniel Naber wrote:
On 2013-09-25 07:38, Ruud Baars wrote:
In the menu: Undo addition (What is ist?)
In the file open dialog: Look in, File Name, Files of type etc.
Are these hardcoded?
Yes, but not by LT
Daniel, this is what I get when trying to start LT:
OpenJDK Runtime Environment (IcedTea6 1.12.6) (6b27-1.12.6-1ubuntu0.12.04.2)
OpenJDK 64-Bit Server VM (build 20.0-b12, mixed mode)
ruud@ruud-desktop:~/Bureaublad/LanguageTool-2.3-SNAPSHOT$ java -jar
languagetool-standalone.jar
Exception in
Daniel, I have been able to get Java working (removed all java-likes,
and installed 7)
But the Dutch menus have not been adjusted.
I redid all of it in Transifex, hoping this time it will be saved ...
Ruud
--
October
Sorry, I still haven mastered coding on the level required, nor any
knowledge about checking in/out.
I will try the snapshot tomorrow.
Ruud
On 24-09-13 20:09, Daniel Naber wrote:
On 2013-09-24 18:31, Ruud Baars wrote:
I redid all of it in Transifex, hoping this time it will be saved
I see them too.
Kubuntu.
Ruud
On 24-09-13 20:22, gulp21 wrote:
does anybody see tooltips when hovering over the icons in the
stand-alone version? They seem to be in the code, but they don't show up
for me.
I see them. LT from 28/08/2013 on Kubuntu 12.04
Regards
Markus
?
Ruud
On 24-09-13 20:43, Ruud Baars wrote:
Sorry, I still haven mastered coding on the level required, nor any
knowledge about checking in/out.
I will try the snapshot tomorrow.
Ruud
On 24-09-13 20:09, Daniel Naber wrote:
On 2013-09-24 18:31, Ruud Baars wrote:
I redid all
I will be starting improving Dutch again after the coming release.
This will include redoing postags all over, and adapting and improving
rules.
Ruud
--
LIMITED TIME SALE - Full Year of Microsoft Training For Just
I tried to translate all untranslated Dutch interface elements using
Transifex.
It is not clear where all options are used, so I just tried to keep it
consistent.
Ruud
--
LIMITED TIME SALE - Full Year of Microsoft
As long as there is a Hunspell one, and it is of good quality.
On 20-07-13 08:51, Daniel Naber wrote:
Am 20.07.2013 00:19, schrieb Daniel Naber:
http://code.google.com/p/zemberek/ [4]
Yes, if the MPL license of the project also applies to the word list
we
could use that.
Actually, the
What about abbreviations?
Hunspell accepts abbreviations without the . even when these are
actually incorrect.
LT knows about abbreviations, using srx for sentence endings at least,
but what about in-sentence abbreviations?
Ruud
On 20-07-13 08:51, Daniel Naber wrote:
Am 20.07.2013 00:19,
By the way, I could help with words frequencies for some langauges.
e.g. Portuguese, Spanish, Dutch.
Ruud
On 16-07-13 14:20, R.J. Baars wrote:
Coding word frequencies as a character is fine. I think it would be
classes, logarithmic as far as I am concerned.
Ruud
W dniu 2013-07-16 00:03,
Might a change to the spacebefore-detection be an option?
Like specifying space type instead of just No or Yes ?
Ruud
On 23-05-13 12:58, Marcin Miłkowski wrote:
W dniu 2013-05-23 11:32, Nathan Wells pisze:
So I just confirmed with some farther testing that spacebefore considers
a zero-width
Would this touch all languages?
In which way then?
For Dutch it is very likely spliting at a dash will end up accepting
wrong words..
Ruud
On 15-05-13 18:04, Daniel Naber wrote:
Am 15.05.2013 05:15, schrieb Andriy Rysin:
As this change touches the core code I wanted to review it here first
Marcin did a change for the Dutch tokenizer to support the - as part of
the token, not as separator.
Ruud
On 15-05-13 05:15, Andriy Rysin wrote:
Hi all
I had some requests for Ukrainian module to support hyphenated words
better in our spellchecker. In general Ukrainian has a lot of special
). The first is ideal for number generation
e.g.; the second more useful for normal words.
Ruud
On 09-05-13 15:57, Marcin Mi?kowski wrote:
W dniu 2013-05-08 19:46, Daniel Naber pisze:
Am 08.05.2013 19:34, schrieb Ruud Baars:
Daniel, maybe it is an idea to get a page somewhere to get the info
Daniel, maybe it is an idea to get a page somewhere to get the info
together on requirements for compounding in the spell checker.
A wiki page?
I have been contributing a lot of ideas to the most recent version of
Hunspell, required things for good Dutch support.
Ruud
On 08-05-13 18:08,
The current hyphenation mechanisms are not able to support Dutch very
well. The number of levels is too little, the possible pattern length
too; there is no concept of compounding, nor of correct changing a word
when hyphenating.
Ruud
On 07-05-13 16:36, Marcin Miłkowski wrote:
Ah, this of
Schreiber wrote:
The problem with the compounds in Hunspell that Ruud described exists
for German as well. Just saying.
Am 03.05.2013 13:07, schrieb Ruud Baars:
Hi.
Finally I have a full keyborad, to elaborate a bit on the expansion issue.
Spell checking is supposed signal any incorrect word
Hi.
Finally I have a full keyborad, to elaborate a bit on the expansion issue.
Spell checking is supposed signal any incorrect word. So most correct
words should be accepted.
There are words in between though. Words that are technically correct,
but in everyday language use mocht commonly a
Since Dutch and German are very much alike, we might tem up on thinking
what we need for that.
I noticed German compounding has been implemented in Hunspell quite
differently from my Dutch solution.
When compouding, Dutch needs:
- border-detection of case and letter combinations, forcing a
For some purpose, I might need a spell checker covering all languages.
Word and expected language in, word and alternatives out, including
languages:
example for Dutch : sex (NL) = sex (EN), seks (NL), sexy (NL/EN)
Maybe this is a way to detect language better then tika does, maybe
detect
You can assume it is a replacement.
I never registred a bug. Where do I do that?
On 11-04-13 16:36, Daniel Naber wrote:
On 10.04.2013, 21:32:19 Ruud Baars wrote:
data.taaltik.nl/LanguageTool/Accepted_by_OpenTaal_210.txt.zip , 3.5 MB
So this is the list of words that should be accepted, isn't
Forget that, I found it.
On 11-04-13 16:41, Ruud Baars wrote:
You can assume it is a replacement.
I never registred a bug. Where do I do that?
On 11-04-13 16:36, Daniel Naber wrote:
On 10.04.2013, 21:32:19 Ruud Baars wrote:
data.taaltik.nl/LanguageTool/Accepted_by_OpenTaal_210.txt.zip
What I could do for the Dutch speller to improve it (it is just too
limited now) is run my millions of words through it, finding out which
ones are accepted, thus effectively detecting all accepted compounds.
How would I from that plain list generate the spelling list, or could I
just post the
There is a relatively fast way to deal with alternatives.
Fore every word one could compute a number, using this recipe:
- lowercase the character
- unaccent the character
- have the ascii-value of the letter, raise it to the fifth power
- add all these numbers
This makes a 'fast lookup'-number
Works fine for me.
Ruud
On 01-04-13 00:34, Daniel Naber wrote:
Hi,
the LanguageTool 2.1 release has now been uploaded to
http://www.languagetool.org/download/?C=M;O=D
It's not yet announced, so please take the time to test it. Unless some
serious problem shows up, I will announce and link
One setback is the spell checker. It has the same issue most
English-based spellcheckers have: suggest a space, splitting up words.
Since the words list is far from complete, it suggests a lot of wrong
split-ups.
Either the word splitting should be disabled, or the words list enhanced
quite a
The university of Nijmegen (The Netherlands) has been working on tool
combinations that do exactly that, use machine learning from large
corpora, Dutch as well as English.
The Dutch front-end is called valkuil.net. More info on the tools and
usage is here:
In case you need it, here is the official translation of Dutch in a
lot of languages:
(Unfortunately, the language it is in, is mentioned in Dutch...)
Ruud
Afrikaans:
Nederlands
Albanees:
of language names in Slovenian is used, so I
hope you are not suggesting to replace the existing Slovenian name
for Dutch with the one on your list?
Lp, m.
2013/2/16 Ruud Baars baar...@xs4all.nl
In case you need it,
here is the official
If regexp is good enough, you could do that in disambiguation. But no
certainty that the word as a whole is correct.
Ruud
On 06-02-13 13:05, Jaume Ortolà i Font wrote:
2013/2/6 Ruud Baars baar...@xs4all.nl mailto:baar...@xs4all.nl
Jaume,
I was not working on compouding; that has
Naber wrote:
On 06.02.2013, 08:05:16 Ruud Baars wrote:
When I know a bit more, I could try to adjust the prototype code to
support multiple languages by design.
jwordsplitter can be extended to support languages other than German. You
need to implement a class that extends
, 19:41:26 Ruud Baars wrote:
I have had a look at that code a long time ago. I found it hard to
understand.
The most recent code on github (not in the latest release yet) has been
rewritten and should be easier to follow. Help for checking out the code is
on github (basically just git clone
A long time ago I prototyped a word uncompounder for Dutch.
Though it worked, it was far from elegant and supporting only Dutch.
Earlier this week I found a more elegant soution, able to uncompound
words like
'langetermijnplanning' into 'lange termijn planning'.
In Dutch there are 4 possible
+1
On 25-01-13 18:01, Dominique Pellé wrote:
Hi
The directory tree of LanguageTool source and resource files is quite
deep.
I understand that this is mostly due to Java conventions. But now
that languages
are separated in modules, I don't think we need to have 2 directories
with the name
As far as I am concerned, I think it is very technical.
I will work from the nightly builds.
On 24-01-13 13:15, Jaume Ortolà i Font wrote:
2013/1/23 Daniel Naber list2...@danielnaber.de
mailto:list2...@danielnaber.de
Don't hesitate to ask if you have questions.
Hi Daniel,
I miss an
Common errors in Dutch are:
kado where cadeau is right : a k-c switch, combined with a eau-o switch.
2 replacements in one word.
To be honest, Hunspell applies only 1.
A differnet example is the combination stofgezogen-gestofzuigd.
These are all not typo's, but corrections are desired
If suggestions are listed in a suggestion list, why having suggestions
in the message at all?
There is a short message, a long message with suggestions, and then a
suggestion list.
Maybe it is possible to reduce the messages to 1, and leave the
suggestion out.
Ruud
On 16-01-13 18:44, Daniel
Daniel,
I am working on just that for Dutch.
Ruud
On 17-01-13 00:35, Daniel Naber wrote:
Now that we have a nice online check, we might contact website owners who
offer automatic spell or grammar checking, asking them if they would like to
integrate LT into their service.
For example,
I am cosidering this wild idea:
- ignore tradtional spell checking of just single words
- start spellchecking word groups from 1 to n (3 to start with) words.
- use an engine to suggest known correct alternatives for known
erroneous word groups
- suggest more common much alike word groups.
A
-12 21:30, Ruud Baars wrote:
Thanks.
Ruud
On 30-12-12 21:23, Daniel Naber wrote:
On 29.12.2012, 21:28:22 Ruud Baars wrote:
I added land-in-waarts+ to compounds.txt.
The very latest snapshot at http://www.languagetool.org/download/snapshots
now contains a fix for that.
Regards
Daniel
unless you add a disambiguator, this is not useful to you anyway.
To learn more about the disambiguator in LT, see:
http://wiki.languagetool.org/developing-a-disambiguator
Regards
-- Dominique
Ruud Baars baar...@xs4all.nl mailto:baar...@xs4all.nl wrote:
Could you make clear what
Is this rule case sensitive? That is relevant sometimes (Aids Fonds
[=organisation] vs. aidsfonds [ any fund])
On 30-12-12 21:30, Ruud Baars wrote:
Thanks.
Ruud
On 30-12-12 21:23, Daniel Naber wrote:
On 29.12.2012, 21:28:22 Ruud Baars wrote:
I added land-in-waarts+ to compounds.txt
Could you add case sensitivity on the wish list?
Ruud
On 31-12-12 16:01, Daniel Naber wrote:
On 31.12.2012, 15:51:37 Ruud Baars wrote:
Is this rule case sensitive? That is relevant sometimes (Aids Fonds
[=organisation] vs. aidsfonds [ any fund])
It's currently case-insensitive.
Regards
Or more plain, add the exclamation mark at the start for case-sensitivity.
Ruud
On 31-12-12 16:18, Daniel Naber wrote:
On 31.12.2012, 16:12:18 Ruud Baars wrote:
Could you add case sensitivity on the wish list?
It's already on the wish list for a long time, as it's also needed (rarely
Please don't lock thing to Eclipse. Eclipse is a tough thing to learn
fro non-developers like me.
I would rather stick to my plain text editor to edit rules and check
them from the command line.
Ruud
On 30-12-12 23:01, Mauro Condarelli wrote:
Just my 2c...
IMHO there's no gain (aside from
I started LT GUI as a server, and was using that.
In the GUI, I disabled a rule using the link.
But after that, the rul was still applied by the server.
Ruud
--
Master Visual Studio, SharePoint, SQL, ASP.NET, C# 2012,
Could the rule that uses compounds.txt be adjusted to do a suggestion?
Ruud
--
Master Visual Studio, SharePoint, SQL, ASP.NET, C# 2012, HTML5, CSS,
MVC, Windows 8 Apps, JavaScript and much more. Keep your skills current
regels - klik aan om weer in te schakelen: Waarschijnlijk
een spelfout
Klaar, 1 aandachtspunten
Could it be a string translation issue in the where I missed a parameter?
Ruud
On 29-12-12 20:12, Daniel Naber wrote:
On 29.12.2012, 20:04:36 Ruud Baars wrote:
Could the rule that uses
There apparently is no rule checking for tv , e double quote, caught
between 2 letters. A single qout is possible in Dutch, a double never.
Is this true for all languages? Would a rule for this a small change in
one of the java rules, or should I make a xml rule for this?
Ruud
There is a problem with Hunspell. It looks like you have not implemented
the latest version.
The call from LT for the word 'wulfsite' results in the erroneous
suggestion 'wulf site'. This is quite serious for Dutch!
Using the incorporated .dic and .aff and entering the same word, results
in
I hit an example that is quite strange:
Een week geleden zei de brancheorganisatie rekening te houden met een
extra omzet van 230 miljoen euro in de dagen voor Pasen. ,Die
inschatting lijkt aan de voorzichtige kant te zijn geweest. Vermoedelijk
komt dat bedrag hoger uit.
It is quite clear
How do I switch of the spell-checking rule in the GUI/options?
On 28-12-12 13:48, Ruud Baars wrote:
There is a problem with Hunspell. It looks like you have not implemented
the latest version.
The call from LT for the word 'wulfsite' results in the erroneous
suggestion 'wulf site
This is not 2 single quotes, but 1 double. The error there is no space,
where there should be one, left or right.
Ruud
On 28-12-12 18:44, Marcin Miłkowski wrote:
W dniu 2012-12-28 13:39, Ruud Baars pisze:
There apparently is no rule checking for tv , e double quote, caught
between 2 letters
both.
Ruud
On 27-12-12 19:53, Dominique Pellé wrote:
Ruud Baars baar...@xs4all.nl mailto:baar...@xs4all.nl wrote:
You don't have to use compounding.
add a flag for the 'affix' 're' etc.
Like
SFX X J 2
SFX X e .
SFX X ere .
(or something better)
and add numbers
But what is offset then?
Ruud
On 27-12-12 22:26, Dominique Pellé wrote:
Ruud Baars baar...@xs4all.nl mailto:baar...@xs4all.nl wrote:
Can anyone please explain what the added value is of the api
values of:
contextoffset and offset
in:
error fromy=0 fromx=0 toy=0 tox=5
Ah, I found the add-on-balk (which is off by default) You would need
that on for the icon to be visible.
Ruud
--
LogMeIn Rescue: Anywhere, Anytime Remote support for IT. Free Trial
Remotely access PCs and mobile devices
The texts are very detailed, and might be tailored for Windows, since I
cannot see FF-bars anywhere in my FF on Kubuntu.
Some terminology should be following Mozilla terminology, which in
unfamiliar to me.
So the translation is there mostly, but not all of it, and certainly not
Reduce it to LT ?
On 22-11-12 16:11, Daniel Naber wrote:
Hi,
does anybody have a problem with the [Languagetool] prefix removed from
the mails on this list? I think there are other ways to filter mails in all
mail clients and it just takes up space better used for the subject.
Regards
Number of reported errors (rules) will structurally grow; length of text
is out of our field of influence. So there is only speeding things up left.
Apart from profiling the code, one could consider profiling the
individual rules and the amount of found hits, related to the
seriousness of the
I am looking for a tool to dump (paragraph data of articles in) a
Wikipedia dump to play utf8 text.
Any advice?
Ruud
--
Monitor your physical, virtual and cloud infrastructure from a single
web console. Get in-depth
In Dutch, ther eis a lot of word confusion.
An example:
paardenwagen = a vehicle (wagen) to transport horses (paarden)
paard-en-wagen = a wagon and horse combination
It is possible to make ruls for these kind of confusions, but neither of
the two is really wrong. One would just want to inform
Daniel, that won't do. Dutch, even with spelling errors, should be
distinquished from Afrikaans.
I'll keep you informed.
On 17-11-12 12:36, Daniel Naber wrote:
On 17.11.2012, 07:49:29 Ruud Baars wrote:
I need to find out the quality of distinction possible between
old-fashiond Dutch
Is there any info on how that works? It is not in the wiki.
Ruud
On 17-11-12 15:42, Daniel Naber wrote:
On 17.11.2012, 12:05:14 Ruud Baars wrote:
Would it be and idea to add A function for this, a bit like
compoundrule, having its own data file?
We already have de/wrongWordInContext.txt
right from the start.
So I am curious how it is trained and used.
Ruud
On 17-11-12 18:05, Susana Sotelo Docio wrote:
Ruud Baars escribiu:
Hi I tried to read the documentation, but that is very technical. Not a
word about what it is really able to, and how it is trained.
Would you know where I
Is LT using language detection from :
https://code.google.com/p/language-detection/
Or something else?
I am currently experimenting with this Google stuff, but the profiles
for anything else then English seem to be relatively poor.
Any opinions on this?
Ruud
, German, Afrikaans, Frysian etc., for better
filtering of a corpus.
Ruud
On 17-11-12 00:16, Daniel Naber wrote:
On 16.11.2012, 18:14:50 Ruud Baars wrote:
Is LT using language detection from :
https://code.google.com/p/language-detection/
We're using http://tika.apache.org
Regards
Daniel
Daniel,
I once started the server by starting the gui from the commandline,
after having the server switched on.
By testen every once and a while, i was able to kille the process
entirely, and start the gui (and server again).
Problem is to let the port be freed entirely.
Maybe it is of help?
Does anyone have a nice solution for multi-word problems like
/inter net/ instead of /internet/ (This is nice covered by the compoundRule)
/wordproblem/ instead of /word problem/ I know spellchecking can do
that, but it is dangerous to have a space suggested in compounding
languages like
Daniel.
For Dutch, the rule 813 is ok; it is the sentence splitter that needs an
exception for these incorrrect abbreviations. It is currently tuned for
only correct abbreviations.
)./token
rule break=no
beforebreak\b(mm|cm|km|mg|kg|h|kW|mW)\.\s/beforebreak
Done.
On 11-11-12 12:40, Daniel Naber wrote:
On 11.11.2012, 07:03:08 Ruud Baars wrote:
The * is good, but indeed; we still miss the differnce like Jan
mentioned.
I see - could you open a bug report / feature request for that?
Regards
Daniel
Daniel, others
I added to compound.txt:
van-waarde-verklaring+
Having /van waarde verklaring/ leads to the suggestion of writing things
together;
/vanwaarde verklaring/ does not, however, nor does / van waardeverklaring/.
That is by design?
Luckily, most of these are reported by the spell
Daniel, would you be willing to think about addint the opposite the + at
the end:
# + at the end of the line will turn off the suggestion that
# uses a hyphen
In some case we would want to only suggest the option with the hyphen:
# - at the end of the line will turn off the suggestion that
#
The * is good, but indeed; we still miss the differnce like Jan mentioned.
I don't know the impact, but it would be nice to have.
Ruud
On 11-11-12 01:35, Jan Schreiber wrote:
Daniel Naber wrote:
On 10.11.2012, 18:37:36 Ruud Baars wrote:
In some case we would want to only suggest
Both tools require or request a local server.
Is that caused by not having (scalable) capacity?
Any idea what would be needed for a central server (farm).
Ruud
On 09-11-12 12:15, Daniel Naber wrote:
On 09.11.2012, 11:00:54 Marco A.G.Pinto wrote:
What I would like to have would be a
Then the translations might be:
Koppeltekenprobleem
Woorden die aaneen horen met koppeltekens, bijvoorbeeld 'zee-egel'
i.p.v. 'zee egel'.
Ruud
On 09-11-12 19:53, Daniel Naber wrote:
On 09.11.2012, 19:37:53 Ruud Baars wrote:
Hyphenation problem
A short error description used in LO/OO's
Guys,
How do I setup the software to generate postag dictionaries?
Marcin once handed me a set, but that got lost over the years, changing
computers etc.
Ruud
--
Don't let slow site performance ruin your business.
Though the input for fsa surely has 13-jarig as a base word for
13-jarige, when I look at the tag output of TL, is see jarig used as the
base word for 13-jarige.
That is not exact. It is close, but not correct when using the relations
to suggest alternatives using the postag. (13-jarig would
Watch the difference between 2 examples, very much alike, apart from the
number-letter difference:
16-jarige[jarig/AJe]
17e-eeuwse[17e-eeuws/AJe]
Ruud
On 14-10-12 21:44, Ruud Baars wrote:
Though the input for fsa surely has 13-jarig as a base word for
13-jarige, when I look at the tag output
Marcin,
For Dutch, tokenisation on - would be really wrong, since it is a real
word character, sometime required, sometimes optional.
Dutch has the phenomenon of 'klinkerbotsing' (sonant collision) when two
single sonants get glued together, and can be mistaken for one of the
two-charcter
There is no Oracle Java package available.
But adding the libreoffice-java-common solved it.
Ruud
Op 30-06-12 16:47, Daniel Naber schreef:
On Samstag, 30. Juni 2012, Ruud Baars wrote:
I also get a java error when trying to install it.
Ubuntu, installed default Java runtime yesterday.
What
How do I invoke this rule from XML or otherwise (not java code)
Ruud
Op 29-06-12 15:53, Marcin Miłkowski schreef:
Hi all,
shouldn't we make also SimpleReplaceRule a spelling rule and display its
results underlined in red?
Regards,
Marcin
When words get longer, edit distance should increase too.
A % maybe.
In the matter of 'rep', the same issue is present. 1 rep is not enough:
e.g. the k/c mutation for dutch : kokos vs cocos.
Multiple replacements in one word could be needed.
(It is a change request for Hunspell too).
Also.
There is a words list for Dutch at www.opentaal.org. Relatively small, just
450.000 words, including proper names, but it is the best open one there is.
Ruud
Marcin Miłkowski list-addr...@wp.plschreef:
Hi all,
considering that hunspell is so slow and that for many languages, we
already have
Aspell for debian is generated from the same source. But other aspell versions
are not.
Better use the real, officially aproved list; it has te certificate of the
Dutch language union; the govt in fact!
Marcin Miłkowski list-addr...@wp.plschreef:
Ruud,
what about aspell-nl? Is that the same?
, I might be the only one using it as a service of
from the command line, later this month.
Ruud
Marcin
W dniu 2012-06-22 14:05, Ruud Baars pisze:
Aspell for debian is generated from the same source. But other aspell
versions are not.
Better use the real, officially aproved list; it has te
Dominiqu,
Marcin once did that trick for Dutch, allowing for single quotes within
a word.
Might check that part of the code.
Dutch even has the /Jans'/ (owned by Jans) as a word and /'s morgens/;
in both cases these are not quotes, but apostrophes.
Ruud
Op 22-06-12 20:31, Dominique Pellé
This might also be different for words rigjht in the Hunspell
dictionary, versus words formed by affixes, and more so, words
recognized using compounding.
So be careful with these results ...
Ruud
Op 22-06-12 21:46, Marcin Miłkowski schreef:
W dniu 2012-06-22 20:31, Dominique Pellé pisze:
I am willing to help starting a better dic. Time is limited though.
Since we are both in the Netherlands,(?), getting together is the fast option.
Ruud
Marcin Miłkowski list-addr...@wp.plschreef:
W dniu 2012-06-04 22:51, Dominique Pellé pisze:
Hi
Another problem with spell checking, is the
For Europe, one should use British.
For worldwide: US.
I think this is a gov viewpoint, at least in Holland
Daniel Naber list2...@danielnaber.deschreef:
On Sonntag, 3. Juni 2012, Marcin Miłkowski wrote:
I only added de-DE dictionary (frami version) because adding Swiss and
Austrian
...@gmail.com
napisał:
There is no command line for hunspell library. It's not executed as
standalone program.
04-06-2012 11:44 użytkownik Ruud Baars baar...@xs4all.nl napisał:
Command line hunspell needs utf8 to be specified.
Ruud
Marcin Miłkowski list-addr...@wp.plschreef:
W dniu 2012-06
1 - 100 of 118 matches
Mail list logo
