Difference between revisions of "format of Gismu Lists"

From Lojban
Jump to navigation Jump to search
m
 
m (Conversion script moved page Format of Gismu Lists to format of Gismu Lists: Converting page titles to lowercase)
 
(4 intermediate revisions by one other user not shown)
Line 1: Line 1:
  
You will mainly work with a Unicode text file which will probably become a module for Parallel 2 - created by you.
+
The Lojban gismu list is a plain-text ASCII columnar format file.
  
The format for modules is described in the ''Readme.txt'' from Parallel, so I won't repeat it here.
+
* Column 0 is the first character position (a.k.a. column) in each line.
 +
* Columns 1-5 is the gismu itself.
  
How to prepare sound files for Parallel? They must be audible but not too large in size, in order to be downloadable by most users over the Internet. In addition, there is a reason to listen to low-quality sound. Your ears and brain get accustomed to harsh conditions. No pain - no gain! So saving in MP3 format is recommended, with following settings:
+
* Columns 6-19 are up to three assigned rafsi, with position in the field indicating the form CVC (columns 7-9), CCV (columns 11-13) or CVV (columns 15-18).
 +
* Columns 20-39 is the English keyword - a unique word useful for remembering the word but not necessarily a definition; being unique was more critical than being definitional.
  
* Mono
+
* Columns 41-60 is a hint word enclosed in single quotation marks, which may be an alternate word with related meaning that may be used as a memory aid (these might not be unique or unambiguous).
* 16 kBit/s
+
* Columns 62-157 is the official place structure definition.
  
* 8,000 Hz
+
* Columns 159-160 is a letter and number code associating the word with the [[original Lojban textbook|original Lojban textbook]] outline to give the intended teaching order.
 +
* Columns 161-164 is a frequency count from 1993, giving an ordering based on usage up to that point.
  
For recording and editing, I use 44100 Hz resolution.
+
The latter two were used to generate non-alphabetical sorted lists based on
  
After recording the sentence, I cut the initial silence, then cut the final silence, but leaving about 0.1 s to be consumed by MP3 encoder.
+
two plausible orderings in which one might want to learn the words.
  
After that I normalize the sound with Normalize RMS to -10 dB (speech) (Sound Forge feature).
+
* Columns 165+ contain additional information that did not fit into the place structure field, generally explanatory, with cross-references to related words.
 +
 
 +
Originally, only the first three fields were baselined, but the community
 +
 
 +
has treated the entire document as frozen for several years. The [[baupla fuzykamni yfy|baupla fuzykamni yfy]] is
 +
 
 +
responsible for making any changes to it at this point.
 +
 
 +
For information on the correct interpretation of natlang text in the gismu lists (columns 62-157 and 165+), see [[BPFK gismu Section: Parenthetical Remarks in Brivla Definitions]].
 +
 
 +
[[User:Bob LeChevalier|Bob LeChevalier]]

Latest revision as of 08:15, 30 June 2014

The Lojban gismu list is a plain-text ASCII columnar format file.

  • Column 0 is the first character position (a.k.a. column) in each line.
  • Columns 1-5 is the gismu itself.
  • Columns 6-19 are up to three assigned rafsi, with position in the field indicating the form CVC (columns 7-9), CCV (columns 11-13) or CVV (columns 15-18).
  • Columns 20-39 is the English keyword - a unique word useful for remembering the word but not necessarily a definition; being unique was more critical than being definitional.
  • Columns 41-60 is a hint word enclosed in single quotation marks, which may be an alternate word with related meaning that may be used as a memory aid (these might not be unique or unambiguous).
  • Columns 62-157 is the official place structure definition.
  • Columns 159-160 is a letter and number code associating the word with the original Lojban textbook outline to give the intended teaching order.
  • Columns 161-164 is a frequency count from 1993, giving an ordering based on usage up to that point.

The latter two were used to generate non-alphabetical sorted lists based on

two plausible orderings in which one might want to learn the words.

  • Columns 165+ contain additional information that did not fit into the place structure field, generally explanatory, with cross-references to related words.

Originally, only the first three fields were baselined, but the community

has treated the entire document as frozen for several years. The baupla fuzykamni yfy is

responsible for making any changes to it at this point.

For information on the correct interpretation of natlang text in the gismu lists (columns 62-157 and 165+), see BPFK gismu Section: Parenthetical Remarks in Brivla Definitions.

Bob LeChevalier