Laboratoire ICAR » Plateforme TXM

Currently, word selection in editions is based on browser character selection.

TXM is a word based text analysis tool, not character.

For word based hypertext links, selection feedback should be based on word boundaries.

• a) manage a character selection + word selection
  • the user can see/compare/verify his usual character selection
  • depending on context, tools can use a fine character selection or a bigger word selection
• b) manage only a word selection, two possibilites
  • NOW b.1) use character selection to represent word selection
  • LATER b.2) use any browser character style to represent word selection
    • example: frame rectangle in BRAT http://brat.nlplab.org/img/examples/PMID-20300060-full.png

• single left click = word selection
• double left click -> remove word boundaries character selection (depends on writing system and its implementation)
• span left selection -> expand to underneath words selection

• single right click
  • if word selection call command on selected words
  • if no word selected call command on underneath word
    • NOW a) select the word before calling the command
    • LATER b) don't select the word before calling the command