| Ticket UUID: | 795032 | |||
| Title: | [package require uri] alter some vars on global level | |||
| Type: | Bug | Version: | None | |
| Submitter: | pointsman | Created on: | 2003-08-26 00:45:20 | |
| Subsystem: | uri | Assigned To: | andreas_kupries | |
| Priority: | 5 Medium | Severity: | ||
| Status: | Closed | Last Modified: | 2003-08-28 04:27:54 | |
| Resolution: | Fixed | Closed By: | patthoyts | |
| Closed on: | 2003-08-27 21:27:54 | |||
| Description: |
This script: set host test package require uri puts $host returns: ((([A-Za-z0-9][A-Za-z0-9-]*[A-Za-z0-9]|[A-Za-z0-9])\.)*([a-zA-Z][A-Za-z0-9-]*[A-Za-z0-9]|[a-zA-Z]))|([0-9]+\.[0-9]+\.[0-9]+\.[0-9]+)) It's not, that [package require uri] simply pollutes the global namespace. A package require uri puts $host gives "can't read "host": no such variable" as expected. The var 'host' must be set prior to the [package require uri] to be altered by the package request. The same problem shows with other variable names, for example 'escape', 'hostOrPort', 'uChar' and very probably some more. As far, as I've tracked it down, it seems, that proc ::uri::register or the script args of the various ::uri::register calls at the end of uri.tcl are the source of the problem. | |||
| User Comments: |
patthoyts added on 2003-08-28 04:27:54:
Logged In: YES user_id=202636 Its the various set statements. If the given variable exists in the global namespace then that is set. Remember that namespace name resolution looks at the current namespace and the global namespace only. The fix is to use variable everywhere -- this passes the test suite and doesn't affect existing global variables. | |||
Home
Blog
Docs
Downloads
Tickets
Timeline
Branches
Tags
Wiki