I have a schema definition in a XSD file which is provided by ISO20022. This schema will need to be used in a swagger/openAPI definition (in yaml format). Since the XSD file has about 1000 lines, manual work is impracticable. This old thread mention some solution, but it is not straightforward.
Does anyone know any tool which provides an easy way to import the schema definitions from a XSD file into a swagger/openAPI yaml file?
You could try xsd2json from the npm module jgexml. It was written to do precisely this for a large API specified in XSD.
I could not escape from manual work in this task. What I've done was using "xsd2json" to convert the XSD schema to JSON. Then, I used the website www.json2yaml.com to get it as YAML. Afterwards, I created a swagger file myself, then merged the YAML file into it.
Thanks for your responses!
Related
recently i've been asked to implement config files for my system, config for each environment.
When i wanted to use the config i noticed that i dont have it typed, at least not in easy way..
So i created an index file which import and export the config adding an interface to it.
I wonder if i can add a type to my config (somehow) which will force the developers to stick to it and also provide us a type at compilation time.
Its like to have a config.ts file instead of config.json (maybe that what i should do?)
Thanks!
Yes you can - use json schema for this purpose.
To ensure your develpers have not messed up the data - use one of many available validators (i find this one to work pretty nice) to verify json object you read from config file.
Some IDE (vscode for example) support json schemas and can validate it on the fly.
You can have your config stored in *.ts file as exported const variable. Then after requiring it verify it with validator. Unfortunately this approach will not give you errors at compilation type.
I have a lot XSD which access each other.
Since I'm changing to JAXB I'm currently looking for a straightforward way to add namespaces to the xsds according to their folder structure.
My main issue at the moment is to add the namespace to the XSD itself, to the import in the accessing file and also to the prefix definition in the accessing file.
Here a small example (not quite real live)
User.xsd - targetNamespace="common.user"
Message.xsd - targetNamespace="common.message"
Email.xsd - targetNamespace="email" xmlns:user="common.user" xmlns:message="common.message"
import namespace="common.user" schemaLocation="./common/user.xsd"
import namespace="common.message" schemaLocation="./common/message.xsd"
When I now have a new Message.xsd schema I have to duplicate 90% of my xsd header.
I was creating a Namespace.xsd xmlns:user="common.user" xmlns:message="common.message" which is then included by Email.xsd. But accessing e.g. user:name did not work.
Is there a way to save the namespace-prefix definition in a central XSD-file so I do not have to define them in every single xsd?
Also, is there a way to not need to set the namespace in the import when it is already defines in the imported xsd as targetNamespace?
I think (not 100% sure though) that including into the schema A a schema B which imports another schema C will give you access to C in A.
However you will still need to declare the namespace prefix (like xmlns:user="common.user").
ps. Just a warning - never ever do chameleon schemas. It does not seem that you plan to, but i still wanted to warn.
I have a project that does JAXB generation with framework.xsd. This generates a jar with the xsd and the jaxb objects and other classes around that stuff.
Then another group(two different groups) will be extending framework.xsd and subxmling using the schema extends stuff to extend objects in framework.xsd. They also want to generate jaxb objects BUT they want their SomeClass.java to obviously extend my Framework.java and don't want to end up with a whole new heirarchy.
Is this even possible?
How to do something like this? as the solution would need to
tell the jaxb compiler that the namespace yy is already generated so do not generate
tell the jaxb compiler that it needs to refer to the classes in the package zzzzzz or to look at the xjb file from the framework jar file or something.
Is this possible?
thanks,
Dean
You want to use an episode file : http://weblogs.java.net/blog/kohsuke/archive/2006/09/separate_compil.html when generating JAXB classes for your first schema.
$ xjc -episode framework.episode framework.xsd
Then the other group that consumes your framework.jar should:
1) import your schema in their own schema e.g.:
<xsd:import namespace="http://www.myorg.com/framework" schemaLocation="framework.xsd"/>
2) generate their JAXB classes
$ xjc extend.xsd -b framework.episode
(they'll need a copy of your xsd and episode file at xjc time, as well as the framework.jar in the classpath)
Note that according to the blog post above, you can also place the framework.episode file inside your jar (e.g. /META-INF/sun-jaxb.episode for JAXB RI at least - other JAXB impl may have other ways of accomplishing the same thing), so that the -b framework.episode option can be omitted. I personally find it a bit impractical, you still need the XSD anyway.
Is there an XQuery command to load a text file?
I can load an xml document by doing the following;
declare variable $text := doc("test.xml");
But it only seems to work if test.xml is a well-formed xml document. What I want is to load a plain test.txt file into a string variable. Something like this;
declare variable $str as xs:string := fn:loadfile("test.txt");
Can it be done?
I'm using the Saxon engine but I can't find an answer in the saxon documentation.
Indeed you can find one implementation of the file functions in Zorba: http://www.zorba-xquery.com/doc/zorba-1.4.0/zorba/xqdoc/xhtml/www.zorba-xquery.com_modules_file.html
XQuery 3.0 has the function fn:unparsed-text (which was originally defined in XSLT), which does exactly what you want. XQuery 3.0 is still a work in progress though, but whilst there are not many XQuery 3.0 processors available, many XQuery processors already support this function (including Saxon).
There is a standardization effort for this on EXPath. A spec already exists for an XQuery File module that is capable of doing what you describe: EXPath File Module Spec.
Yet, I don't know how many implementations are out there. Saxon doesn't seem to implement it unfortunately (Or, please point me to it). An example implementation is shipped with zorba (see XQDoc Site of Zorba). If you want to know how to get started with zorba, you can check out this tutorial: Get Started with XQuery and Zorba.
XQuery by default( means fn: namespace ) doesn;t have any file-access methods.
MarkLogic :
xdmp:filesystem-file()
xdmp:filesystem-directory()
Zorba:
already mentioned by user457056
Exist
Exist File Module
Saxon since version 9.2 has an extension of fn:collection that can be used to read unparsed text. Here is an example:
collection('file:///c:/TEMP?select=text.txt;unparsed=yes')
This is described under "Changes in this Release" for 9.2. Apparently it is not mentioned in the function library documentation. However it works well and I have been using it a lot.
I have developed a groovy application. Now it has been required
that for feeding the DB a CSV interface must be provided.
That is, I have to load a csv file, parse it and
insert records into the DB in a transactional way.
The question is if there exists for groovy something
like ostermiller utils (a configurable csv parser).
Thanks in advance,
Luis
Kelly Robinson just wrote a nice blog post about the different possibilities that are available to work with CSV files in Groovy.
Groovy and Java are interoperable. Take a look at the documentation for mixed Java and Groovy applications. Any Java class can easily be used in Groovy with no change (plus you have the groovy syntax). If you are interested in the ostermiller utils to do your CSV parsing, you can use it directly from Groovy.
If the ostermiller library does what you want you can call it directly from Groovy. Just put the necessary jars in your groovy\lib directory and you should be ready to go.