Difference between revisions of "Lint"

From Apertium
Jump to navigation Jump to search
Line 47: Line 47:
<pre><e r="XYZ"></pre>
<pre><e r="XYZ"></pre>


* Time, Number, Ordinals, Dates, URLs, emails, separators, proper_names, al to be standard
* Time, Number, Ordinals, Dates, URLs, emails, separators, proper_names,et al to be standard


* Unwanted(repeated) <b/> tags in entries.
* Unwanted(repeated) <b/> tags in entries.

Revision as of 08:38, 27 May 2016

About

This page contains a log of all the different errors that the lint will be designed to handle. Along with it, as it's development progresses updates will be posted here about the different releases.

Along with this, documentation about the lint's working and technicalities will also be specified here.

Monodix

List of issues in monodixex

  • Redundant Entries
<pardef n="di/e__vblex"><e><par n="liv/e__vblex"/></e></pardef>
  • Maintain consistency in the data present in the <r> tag in pardef entries.
<e><p><l>as</l><r>o<s n="prn"/><s n="tn"/><s n="f"/><s n="pl"/></r></p></e>
<e><p><l>a</l><r>abz<s n="prn"/><s n="tn"/><s n="f"/><s n="sg"/></r></p></e>
  • Paradigms should have 2:"_" instead of 1:"_".
<pardef n="outr/o__prn">
<pardef n="/o_meu__prn">
  • Lexical entries should have appropriate attribute names. Here is an example of an invalid format which is currently accepted.
<e fm="funny"><i>funn</i><par n="funn/y__adj"/></e>
  • Repeated tag entries.
<e><p><l>y</l><r>y<s n="adj"/><s n="adj"/></r></p></e>
  • Repeated entries in the dictionary.
<e lm="house"><i>house</i><par n="house__n"/></e>
<e lm="house"><i>house</i><par n="house__n"/></e>
  • The "cm" entry should have only the comma and nothing else.
<e><re>,</re><p><l></l><r><s n="cm"/></r></p></e> 
  • Recursive paradigms.
<pardef n="up__n"><e><i></i><par n="house__n"/></e></pardef>
<pardef n="house__n"><e><i></i><par n="xyz__n"/></e></pardef>
<pardef n="xyz__n"><e><i></i><par n="up__n"/></e></pardef>
  • Incorrect and multichar attributes
<e r="XYZ">
  • Time, Number, Ordinals, Dates, URLs, emails, separators, proper_names,et al to be standard
  • Unwanted(repeated) tags in entries.
<l>as<b/>minhas<b/></l><r>o<b/>meu<s n="prn"/>

Bidix

  • Bidix should not have sdefs that are not in monodix or lexc file.

Transfer

  1. Issue in :
     <code><clip></code> 
    Problem : Check that the attribute listed in part="" is defined using a <def-attr>.

Modes

Tagger

Others

Lexical selection
  • don't use lrx-proc without the -m option.


Consistency of {Multichar_Symbols/sdefs/LISTs/SETs}
  • This may be larger task chopped to dozens of checks, but IMO source of most problems: