XMIR, a Quick Tour

• @yegor256
• 25 Nov 2022
• comments

Last updated at: 17.04.2025

XMIR is a dialect of XML, which we use to represent a parsed EO object. It is a pretty simple format, which has a few important tricks that I share below in this blog post. You may also want to check our schema: XMIR.xsd (it is also rendered in HTML, which may be more readable for some of you).

Consider this simple EO object that prints "Hello, world!":

# App.
[] > app
  [x] > foo
    QQ.io.stdout > @
      QQ.txt.sprintf *1
        "Hello, %s\n"
        x
  foo > @
    "world!"

If we parse it using the EoSyntax class from eo-parser, we will get this XMIR (or very similar):

<object
 xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"
 dob="2024-12-27T11:00:08"
 ms="98"
 revision="27abe8b"
 time="2025-04-17T09:32:04.455112Z"
 version="0.56.0"
 xsi:noNamespaceSchemaLocation="https://www.eolang.org/xsd/XMIR-0.56.0.xsd">
 <listing># App.
[] &gt; app
  [x] &gt; foo
    QQ.io.stdout &gt; @
      QQ.txt.sprintf *1
        "Hello, %s\n"
        x
  foo &gt; @
    "world!"
</listing>
  <o line="2" name="app" pos="0">
    <o line="3" name="foo" pos="2">
      <o base="∅" line="3" name="x" pos="3"/>
      <o base=".stdout" line="4" name="@" pos="9">
        <o base=".io" line="4" pos="6">
          <o base="QQ" line="4" pos="4"/>
        </o>
        <o base=".sprintf" line="5" pos="12">
          <o base=".txt" line="5" pos="8">
            <o base="QQ" line="5" pos="6"/>
          </o>
          <o base="string" line="6" pos="8">48-65-6C-6C-6F-2C-20-25-73-0A</o>
          <o base="tuple" line="7" pos="8">
            <o base=".empty">
              <o base="tuple"/>
            </o>
            <o base="x" line="7" pos="10"/>
          </o>
        </o>
      </o>
    </o>
    <o base="foo" line="8" name="@" pos="2">
      <o base="string" line="9" pos="4">77-6F-72-6C-64-21</o>
    </o>
  </o>
</object>

The <object> is the root element, it will always be there, with a few mandatory attributes:

• ms is how much time in milliseconds it took to parse the object and generate this XMIR file,
• time is the time in ISO 8601 format when the file was generated,
• version is the version of the parser.

The <listing> element contains the source code of the EO object, which was parsed, without any modifications, “as is.”

Errors and Warnings

The <errors> element may have a list of problems discovered by the parser or any other optimizers, as <error> elements. If there are no errors, the <errors> element should not exist in <object>. For example, it may look like this:

<object>
  [...]
  <errors>
    <error severity="warning" line="3">There is an extra bracket</error>
    <error severity="error" line="12">The object 'x' is not found</error>
    [...]
  </errors>
</object>

The errors with the warning severity may more or less safely be ignored. The errors with the error severity will lead to failures in further compilation and processing. There could also be elements with the critical severity, which must stop the processing of the document immediately.

Sheets

The <sheets> element contains a list of all post-processors that were applied to the document after is parsing. We process our XMIR documents using dozens of XSL stylesheets. That’s why the name of the XML element. You may find something like this over there:

<object>
  [...]
  <sheets>
    <sheet>move-voids-up</sheet>
    <sheet>const-to-dataized</sheet>
    <sheet>stars-to-tuples</sheet>
    <sheet>wrap-method-calls</sheet>
    [...]
  </sheets>
</object>

The names you see in the <sheet> elements are the names of the files. For example, wrap-method-calls represents the [wrap-method-calls.xsl] file in the objectionary/eo GitHub repository.

If no XSL stylesheets are applied to XMIR, the <sheets> element should not exist in <object>.

Metas

There may be an optional element <metas> with a list of <meta> elements. For example, if my source code would have this meta at the 3rd line of the source file:

+alias foo com.example.foo

We would see the following in the XMIR:

<object>
  [...]
  <metas>
    <meta line="3">
      <head>alias</head>
      <tail>foo Q.com.example.foo</tail>
      <part>foo</part>
      <part>Q.com.example.foo</part>
    </meta>
    [...]
  </metas>
</object>

Each <meta> element contains parts of the meta. The <head> contains everything that goes after the + until the first space. The <tail> contains everything after the first space. There could be a number of <part> elements, each of which containing the parts of the <tail> separated by spaces.

Objects

The <object> element must contain only one <o/> element which represents an object being parsed. The <o/> element may have a few optional attributes:

• line and pos are the number of the line where the object was found by the parser and the position in the line;
• name is the name of the object, if the object has it;
• base may refer to object formation that is being copied;
• as is the name of the attribute which current object is bound to during the application

There could be no other attributes.

Special cases

1. The <o/> elements that have nested <o> element with name which value is λ are atoms. Atoms must not have base attribute: ```xml

2. The `<o/>` elements with `base` attribute which value is `∅` are **void** attributes.
Void attributes also must have `name` attribute:
```xml
<o name="foo">
  <o name="bar" base="∅"/>
</o>

<o base="Q.org.eolang.bytes" line="6" pos="8">
  <o>48-65-6C-6C-6F-2C-20-25-73-0A</o>
</o>

Such `name` consists of several parts:
- char `a` (ascii 97) that stands for "auto-generated"
- char `🌵` that is just a pretty character prohibited by EO grammar
- number `104` which is joined line and position of the place where
the object is found.

Such names are unique through entire XMIR.

5. If object is bound to a specific attribute not by name but by position, the
`as` attribute may look like:
```xml
<o base="Q.org.eolang.number" as="α2"/>