collect in cluster/array descriptions

[burrbull]

replaces #177
cc @kossnikita

[kossnikita]

Fail with

_svd: stm32l4x2.svd
LCD:
  RAM_COM6:
    _array:
      S?? : {}

RAM_COM6: fields cannot be collected into S?? array. Please, specify description

[burrbull]

cc @Emilgardis

[kossnikita]

What about #178 (review)?
I don't like that this breaks existing patches that previously worked correctly

[burrbull]

What about #178 (review)? I don't like that this breaks existing patches that previously worked correctly

Sorry missed this comment.
Can't reproduce. Is it stm32-rs repo? Could you retry on last stm32-rs master and this PR?

[burrbull]

Saying about "RAM_COMx". It looks like there are missed high parts of "64-bit" registers.

[kossnikita]

Is it stm32-rs repo?

This is
svdtools/example/stm32l4x2.svd
My example about auto collecting discriptions like S??. Before PR this works.

[burrbull]

Is it stm32-rs repo?

This is svdtools/example/stm32l4x2.svd My example about auto collecting discriptions like S??. Before PR this works.

Thanks. I totally forgot about this example.
I think this is one of special cases where we need to specify description manually. (or maybe even collect should not work at all).
The issue is names of these fields. They are looks like numerical (and can be parsed as numerical) but instead of S0, S1, S2, .., S31 they are S01, S02, S03, .., S31. So before this patch result looks like (both description and dimIndex are incorrect):

        <register>
          <name>RAM_COM6</name>
          <displayName>RAM_COM6</displayName>
          <description>display memory</description>
          <addressOffset>0x44</addressOffset>
          <size>0x20</size>
          <access>read-write</access>
          <resetValue>0x00000000</resetValue>
          <fields>
            <field>
              <dim>32</dim>
              <dimIncrement>0x1</dimIncrement>
              <dimIndex>0-31</dimIndex>
              <name>S%s</name>
              <description>S00</description>
              <bitOffset>0</bitOffset>
              <bitWidth>1</bitWidth>
            </field>
          </fields>
        </register>

After patch it fails on descriptions.
But real issue is it should produce <dimIndex>00,01,02,..,31</dimIndex> instead of <dimIndex>0-31</dimIndex>.

UPD. After last commit it produces:

        <register>
          <name>RAM_COM6</name>
          <displayName>RAM_COM6</displayName>
          <description>display memory</description>
          <addressOffset>0x44</addressOffset>
          <size>0x20</size>
          <access>read-write</access>
          <resetValue>0x00000000</resetValue>
          <fields>
            <field>
              <dim>32</dim>
              <dimIncrement>0x1</dimIncrement>
              <dimIndex>0-31</dimIndex>
              <name>S%s</name>
              <description>S%s</description>
              <bitOffset>0</bitOffset>
              <bitWidth>1</bitWidth>
            </field>
          </fields>
        </register>

but there is still issue with dimIndex which is somewhere in svd-rs or svd-encode.

[kossnikita]

Looks good now

<field>
  <dim>32</dim>
  <dimIncrement>0x1</dimIncrement>
  <dimIndex>00,01,02,03,04,05,06,07,08,09,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31</dimIndex>
  <name>S%s</name>
  <description>S%s</description>
  <bitOffset>0</bitOffset>
  <bitWidth>1</bitWidth>
</field>

[kossnikita]

I have problem with this but I don't understand the problem
Without collecting DT10 works good

  _array:
    DT?,DT[1][0]:
      description: Battery powered domain data register %s
    DT??:
      description: Battery powered domain data register %s

<registers>
    <register>
      <name>DT1</name>
      <displayName>DT1</displayName>
      <description>Battery powered domain data
      register (BPR_DTx)</description>
    </register>
    <register>
      <name>DT2</name>
      <displayName>DT2</displayName>
      <description>Battery powered domain data
      register (BPR_DTx)</description>
    </register>
    <register>
      <name>DT3</name>
      <displayName>DT3</displayName>
      <description>Battery powered domain data
      register (BPR_DTx)</description>
    </register>
    <register>
      <name>DT4</name>
      <displayName>DT4</displayName>
      <description>Battery powered domain data
      register (BPR_DTx)</description>
    </register>
    <register>
      <name>DT5</name>
      <displayName>DT5</displayName>
      <description>Battery powered domain data
      register (BPR_DTx)</description>
    </register>
    <register>
      <name>DT6</name>
      <displayName>DT6</displayName>
      <description>Battery powered domain data
      register (BPR_DTx)</description>
    </register>
    <register>
      <name>DT7</name>
      <displayName>DT7</displayName>
      <description>Battery powered domain data
      register (BPR_DTx)</description>
    </register>
    <register>
      <name>DT8</name>
      <displayName>DT8</displayName>
      <description>Battery powered domain data
      register (BPR_DTx)</description>
    </register>
    <register>
      <name>DT9</name>
      <displayName>DT9</displayName>
      <description>Battery powered domain data
      register (BPR_DTx)</description>
    </register>
    <register>
      <name>DT10</name>
      <displayName>DT10</displayName>
      <description>Battery powered domain data
      register (BPR_DTx)</description>
    </register>
    <register>
      <name>DT11</name>
      <displayName>DT11</displayName>
      <description>Battery powered domain data
      register (BPR_DTx)</description>
    </register>
...
</registers>

    Caused by:
        0: Processing device `AT32F403`
        1: According to `BPR`
        2: Processing peripheral `BPR`
        3: Collecting registers matched to `DT?,DT[1][0]` in array
        4: BPR: registers cannot be collected into DT?,DT[1][0] array. Please, specify displayName

[burrbull]

cc @kossnikita , @Emilgardis looks like solution is founded

[burrbull]

let's merge this

[kossnikita]

Wait a little longer please. I haven't finished my tests yet.

[kossnikita]

LGTM. Could you update changelog?

[burrbull]

LGTM. Could you update changelog?

Sure.