Internal wikidot Regular Expressions

Internal wikidot Regular Expressions

leiger 23 Dec 2010 23:33

I would like to know what regular expressions Wikidot uses for checking the validity of a site or page name.

For site, I think this is okay?

[-a-z0-9]

But for page names, I have no idea which symbols are allowed and which are disallowed without trial and error. Having the exact same regex that is used in the wikidot code would be perfect as then I know it's right.

Reason: checking site/page validity before calling the API. If I can check it myself, things will run a lot faster and it's one less call to the API that results in an error/exception being sent back.

---

~ Leiger - Wikidot Community Admin - Volunteer
Wikidot: Official Documentation | Wikidot Discord server | NEW: Wikiroo, backup tool (in development)

Last edited on 23 Dec 2010 23:33 by leiger Show more

Re: Internal wikidot Regular Expressions

Gabrys 23 Dec 2010 23:58

You can pass (almost) any full page name to API calls. Wikidot will convert it to "valid" name just like it does when making full page name out of title or URL. The general rules for validity of page names are:

• latin letters and digits are allowed [0-9a-z]
• dash (-) is only allowed when it's surrounded by latin letters or digits or mix of those (page-name, page-0-name, 10-01), two dashes in row, dash in the beginning or in the end are not allowed
• empty name is not allowed

The same rules applies to category names. Category name and page name in full page name is separated by ":". Full page name can have no category in which case ":" is omitted too. These pages belong to category named "_default".

Additional rules:

• "_" is allowed in the front of full page name when there's no ":" in page name
• "_" is allowed in full page name if it follows ":" and it's followed by a digit or letter
• (note "_" is not allowed in category name)
• there can be many ":"s in the full page name, but each part between ":"s must meet the page name rules. String before the first ":" is category name, string after the first ":" (including the remaining ":"s) is page name
• as a special exception the full name <category>:_ IS allowed but discouraged

All these rules can be checked at once using one regular expression, but I think it is more useful to explain them in human language.

Of course rules for constructing valid full page name from user-entered string are even more complicated, but they work quite as you would expect, so I will omit them now.

---

Piotr Gabryjeluk
visit my blog

Re: Internal wikidot Regular Expressions

Gabrys 24 Dec 2010 00:05

For site, it's [-0-9a-z] with the following additional rules:

• only one dash is allowed in a row
• dash can't be located at the beginning or the end of the name
• site name must be at least 3 characters long (with exception for wikidot internal sites/domains)

Summing up, the regexp for site names would be something like this:

[0-9a-z]+(-[0-9a-z]+)+|[0-9a-z]{3,}

---

Piotr Gabryjeluk
visit my blog

Last edited on 24 Dec 2010 00:12 by Gabrys Show more

Re: Internal wikidot Regular Expressions

leiger 24 Dec 2010 01:55

Thanks! Much appreciated!

Seems I have made a mistake already with this! :) In STE 3.22 if you use the "input-page" variable, it asks the user for a page name and that is only allowed to have 0 or 1 occurrences of ":". You've written above that multiple occurrences of ":" are allowed so I will need to change that.

Mainly I needed a way to ensure that the page name and site names are valid, but without having to make a call to the API to do the checking/validation. These rules are exactly what I needed!

---

~ Leiger - Wikidot Community Admin - Volunteer
Wikidot: Official Documentation | Wikidot Discord server | NEW: Wikiroo, backup tool (in development)

Last edited on 24 Dec 2010 02:05 by leiger Show more