-
I've made a fix for this problem. Not thorough testing but seems to work.
--- D:/sdk/kwalify-0.7.1/lib/kwalify/parser/yaml.rb Mon Jan 28 08:57:46 2008
+++ c:/Ruby/lib/ruby/gems/1.8/gems/kwalify-0.7.1/lib/kwalify/parser/yaml.rb Mon Nov 09 14:14:50 2009
@@ -233,11 +233,11 @@
def parse_block_value(level, rule, path, uniq_table, container)
skip_spaces_and_comments()
## nil
-...
2009-11-09 13:37:01 UTC by mikaelpahmp
-
Here's a test for the Validator:
public void testMappingWithStringInKey() throws Exception {
_schema = "" +
"type: map\n" +
"mapping:\n" +
" 'foo bar':\n" +
" type: str\n";
_valid = "'foo bar': value";
_invalid = "foobar: value";
_error = ":key_undefined : (line 1)[/foobar]...
2009-09-01 17:44:30 UTC by sixty4bit
-
The parser does not allow spaces in the key
foo bar: value
is considered invalid. I have updated PlainYamlParser to allow spaces.
2009-09-01 17:42:50 UTC by sixty4bit
-
for the 'map' type there doesn't seem to be a way to restrict key names; example:
type: map
pattern: /^[0-9]+$/
mappings:
"stuff":
2009-04-03 02:59:54 UTC by nobody
-
irb> Kwalify::Yaml.load({ 'foo' => [ 'bar', 'baz'] }.to_yaml)
Kwalify::SyntaxError: 3:1 [/] document end expected (maybe invalid tab char found).
It seems Kwalify doesn't recognize '- ' as valid indentation - manually entering a YAML-string where
the sequence is indented with spaces before the dash makes it parse just fine.
2009-04-02 09:19:35 UTC by kvs
-
Currently, kwalify cannot support something like the following:
users:
steve:
uid: 501
shell: /bin/bash
james:
uid: 502
shell: /bin/bash
When 'steve' and 'james' are keys that are not defined in the schema. Instead, kwalify would support the following:
users:
- name: steve
uid: 501
shell: /bin/bash
- name: james
uid: 502
shell:
2008-08-21 00:48:55 UTC by nobody
-
The bool type is not documented. From experimenting it seems that the validator accepts values like yes, no, true, false etc.
Regards,
Slaven.
2008-08-11 17:31:00 UTC by eserte
-
The float type is not documented. The scientific notation e.g. "1e4" is not supported, though ruby itself seems to support this notation.
Furthermore, integer values like the value "1" seems to be not a valid float, which I found counterintuitive (as ints are normally part of floats). Therefore I think this type is not very useful, and number should be used instead.
Regards,
Slaven.
2008-08-11 17:29:56 UTC by eserte
-
The "str" type is not documented. By experimenting it seems that "str" is every "text" which is not a "number" (which I find counter-intuitive, and almost never useful --- strings are normally valid to hold integers or numbers, too, in most if not all programming languages)
Regards,
Slaven.
2008-08-11 17:29:08 UTC by nobody
-
The "int" type is not documented. It is not clear if the scientific notation (e.g. 1e4) is valid for ints or not. By experimenting, it seems they are not in the ruby implementation.
Regards,
Slaven.
2008-08-11 17:26:19 UTC by nobody
