Difference between revisions of "Syntax checking/SRS"
(3 intermediate revisions by one other user not shown) | |||
Line 1: | Line 1: | ||
+ | [[Category:Projects]] | ||
+ | [[Category:Editor]] | ||
[[Syntax_checking|back to Syntax checking page]] | [[Syntax_checking|back to Syntax checking page]] | ||
Line 11: | Line 13: | ||
## When editing a incorrect expression: | ## When editing a incorrect expression: | ||
### After each added character | ### After each added character | ||
− | # Provide errors in adequate datastructure to visualisation components | + | # Provide errors in adequate datastructure to visualisation components (replaced by is_correct attribute on token) |
##linenumber | ##linenumber | ||
##type of error | ##type of error | ||
Line 20: | Line 22: | ||
* A new expression is correct until it becomes incorrect. | * A new expression is correct until it becomes incorrect. | ||
* The state of an expression can only be changed when the cursor leaves it. | * The state of an expression can only be changed when the cursor leaves it. | ||
+ | |||
+ | = Parser-to-Visualization-Interface = | ||
+ | # The parser invalidates a TOKEN, by settings its attribute is_correct to false | ||
+ | # The parser can also set TOKEN's error_description attribute to the specific error (or error_type if we don't get description) | ||
+ | # An additional data-structure has to be maintained, to pass error summary information to visualization | ||
= Visualization = | = Visualization = | ||
# Highlighting happens only in code editing view | # Highlighting happens only in code editing view | ||
− | |||
# Visualization component should maintain information on line states (underlined y/n) | # Visualization component should maintain information on line states (underlined y/n) | ||
# Underline syntax errors using a (preferably wavy) red line | # Underline syntax errors using a (preferably wavy) red line |
Latest revision as of 07:12, 15 June 2006
Goal
Provide adequate, but non intrusive feedback to user about syntax errors.
Parser
- Should only parse when neccessary:
- When editing a new or correct expression:
- After typing a point, comma, colon, semi-colon, space, tab or line-break
- After leaving the edited expression (using cursor or mouse)
- When editing a incorrect expression:
- After each added character
- When editing a new or correct expression:
- Provide errors in adequate datastructure to visualisation components (replaced by is_correct attribute on token)
- linenumber
- type of error
- further information
- Whenever the visualisation of an expression should change, the parser should call an event handler provided by visualization components
Note: * A new expression is correct until it becomes incorrect. * The state of an expression can only be changed when the cursor leaves it.
Parser-to-Visualization-Interface
- The parser invalidates a TOKEN, by settings its attribute is_correct to false
- The parser can also set TOKEN's error_description attribute to the specific error (or error_type if we don't get description)
- An additional data-structure has to be maintained, to pass error summary information to visualization
Visualization
- Highlighting happens only in code editing view
- Visualization component should maintain information on line states (underlined y/n)
- Underline syntax errors using a (preferably wavy) red line
- On mouse over, show information on error
- Information window could be clickable to provide further error information
- Errors should be accessible over summarization
Suggestions
- To be implemented after parser and visualization components are feature complete
- Suggestions based on information from parse tree
- Analyzation based on soundex or other algorithm