odanoburu / conllu-mode Goto Github PK

to implement a function to clean all the values of the current column (replacing with -).

• adhere to contributing guidelines
  • https://www.gnu.org/software/emacs/manual/html_node/elisp/Coding-Conventions.html
• make PR

when there are meta tokens making the difference between head and index won't take us to the right token.

1. the symbol that marks the \t is not right after the content of the column
2. the columns should be alignment by sentences, avoiding long spaces just because 1 particular long token in one particular sentence.

currently the conllu--token function will only parse simple tokens.

function is being used incorrectly.

it seems that the conllu-move/conllu-parse don't work well when the library is compiled. this might have to do with the macros being used.

if there are many errors emacs slows down noticeably. is this unavoidable or am I doing something wrong? currently this is mitigated by limiting the number of errors reported by the validate script.

This could remunerate the tokens (and adjust the heads) and add a token line with a number already filled?

We also need commands to

1. new multi-word token
2. new empty nodes

realign sentence after call to conllu-edit-field and conllu-clear-field

by @arademaker's suggestion

long columns such as FEATS could be collapsed for facilitate the analysis of the dependencies.

You wrote:

note that the validate.py demands a two-letter language code to do its validation; conllu-mode tries to infer this from the file name, assuming it follows UD filename convention of starting with such a language code (as in pt_bosque-ud-test.conllu). if your file doesn’t start with a language code, conllu-mode will ask you for a language code; if you don’t supply anything, Flycheck will invoke the validate script with the default value of ud.

Could we have variables to supply default values for:

1. validate vs no validate
2. language

So when I open a file with conllu extension, I would not have to answer any question?

currently I'm using delete-and-extract-region to extract the text to be edited. when using flycheck, this triggers a re-run of checker with an empty field. although this does not slow emacs down because it is done asynchronously, it may report a trivial error, so it's better to not delete the region until its replacement is ready.

CC-BY 4.0? Other?