| Type | Issue | Status | waiting | Date | 27-Oct-2009 13:45 |
|---|---|---|---|---|---|
| Version | alpha 92 | Category | Documentation | Submitted by | meijeru |
| Platform | All | Severity | minor | Priority | normal |
| Summary | Undocumented restrictions on allowed character combinations for words |
|---|---|
| Description |
According to the documentation a word can start with an alphabetic character (never mind the definition of that, but see #1230) and any of the following: !&*+-.<=>?_`|~ Subsequent characters can also be a digit or a single quote. However, some combinations are apparently not allowed: a < or > sign inside a word, a word of more than two < or > signs, the combination <=>, and the combinations << and >> used as set-word. Any combination starting with < and continuing with something legal but different from = or > is taken as the start of a tag. I have not been able to explore exhaustively, but if there is a system behind this, I would like to know. |
| Example code |
>> length? [a<<] == 2 >> type? quote <<< ** error! >> type? quote a< ** error! >> type? quote a> ** error! >> type? quote <<< ** error! >> type? quote <=> ** error! >> type? quote <<: ** error! >> type? <!> == tag! |
| Assigned to | n/a | Fixed in | - | Last Update | 18-Feb-2012 10:51 |
|---|
| Comments | |
|---|---|
|
(0001704)
BrianH 27-Oct-2009 20:15 |
Some words are special-cased by the syntax scanner, but otherwise their characters conflict with other types and can't be used. These special cases need to be documented, and the characters that cause conflicts otherwise need to be removed from the list of standard accepted characters. |
|
(0001717)
Carl 28-Oct-2009 17:55 |
I marked this as waiting for documentation changes.
As BrianH indicates, the matter is complicated by what the lexical analyzer will accept as a word. The < and > are overloaded as tags, so their use for words is a special case (for the comparison operators.) Most of the task is to make doc changes, but I do not rule out some extra checks in the code as well. |
|
(0003110)
BrianH 17-Feb-2011 09:49 |
Here is a set of tested rules for the alpha 110 lexical analyzer of R3 words, at least for the ASCII range.
word-char-prefix: charset "+-." word-char-first: charset ["!&*=?^^_`|~" #"A" - #"Z" #"a" - #"z"] word-char-rest: charset ["!&'*+-.=?^^_`|~" #"0" - #"9" #"A" - #"Z" #"a" - #"z"] word-follow: charset [0 - 32 {"()/;[]{}} 127] word-char: complement union word-follow charset "@" money-char: complement word-follow word-follow-error: charset "#$%,<>\" ; See #537 for details word-follow-arrow: charset [0 - 32 "]" 127] ; See #1317 for why "]" is here word-follow-slash: union word-follow charset "+-.#$%,<>\" ; See #1856 for why +-. is here digits: charset "0123456789" word-rule: [ some "/" [end | and word-follow-slash] | "<" [end | and word-follow-arrow] | a: ["," not digits | "\"] any word-char not "@" b: ( cause-error 'syntax 'invalid reduce ["word" copy/part a b] ) | a: ["<>" | "<=" | ">=" | "<<" | ">>" | ">"] [end | and word-follow | some word-follow-error b: ( cause-error 'syntax 'invalid reduce ["word" copy/part a b] )] | a: [ word-char-prefix opt [word-char-first any word-char-rest] | word-char-first any word-char-rest ] [end | and word-follow | "$" any money-char b: ( cause-error 'syntax 'invalid reduce ["money" copy/part a b] ) | some word-follow-error b: ( cause-error 'syntax 'invalid reduce ["word" copy/part a b] )] ] This is just for the syntax of the word! type. The word-rule will match correct words, and will trigger errors in the specific cases shown above. Note: See #1230, #1477, #1478, #1855 and #1856 for the syntax bugs of the other word types. |
|
(0003220)
Ladislav 18-Feb-2012 10:51 |
See also #1229. |
| Date | User | Field | Action | Change |
|---|---|---|---|---|
| 18-Feb-2012 10:51 | Ladislav | Comment : 0003220 | Added | - |
| 18-Feb-2012 03:00 | BrianH | Comment : 0003110 | Modified | - |
| 29-Mar-2011 11:32 | BrianH | Comment : 0003110 | Modified | - |
| 29-Mar-2011 11:01 | BrianH | Comment : 0003110 | Modified | - |
| 29-Mar-2011 04:25 | BrianH | Comment : 0003110 | Modified | - |
| 29-Mar-2011 04:20 | BrianH | Comment : 0003110 | Modified | - |
| 29-Mar-2011 01:59 | BrianH | Comment : 0003110 | Modified | - |
| 17-Feb-2011 22:22 | BrianH | Comment : 0003110 | Modified | - |
| 17-Feb-2011 10:48 | BrianH | Comment : 0003110 | Modified | - |
| 17-Feb-2011 10:48 | BrianH | Comment : 0003110 | Modified | - |
| 17-Feb-2011 09:49 | BrianH | Comment : 0003110 | Modified | - |
| 17-Feb-2011 09:49 | BrianH | Comment : 0003110 | Added | - |
| 28-Oct-2009 17:55 | carl | Comment : 0001717 | Added | - |
| 28-Oct-2009 17:49 | carl | Status | Modified | reviewed => waiting |
| 27-Oct-2009 20:15 | BrianH | Comment : 0001704 | Added | - |
| 27-Oct-2009 20:12 | BrianH | Type | Modified | Note => Issue |
| 27-Oct-2009 20:12 | BrianH | Status | Modified | submitted => reviewed |
| 27-Oct-2009 13:49 | meijeru | Description | Modified | - |
| 27-Oct-2009 13:48 | meijeru | Code | Modified | - |
| 27-Oct-2009 13:48 | meijeru | Description | Modified | - |
| 27-Oct-2009 13:45 | meijeru | Ticket | Added | - |