# Item type for references

**URL:** <https://discourse.devontechnologies.com/t/item-type-for-references/57943>\
**Category:** DEVONthink\
**Tags:** devonthink3\
**Created:** [August 22, 2020, 3:53am UTC](https://discourse.devontechnologies.com/t/item-type-for-references/57943 "2020-08-22T03:53:44Z")\
**Posts on this page:** 15\
**Page:** 1

<div class="post-metadata">

**Author:** ![moosavy](https://discourse.devontechnologies.com/letter_avatar_proxy/v4/letter/m/57b2e6/32.png) [@moosavy](https://discourse.devontechnologies.com/u/moosavy)\
**Post date:** [August 22, 2020, 3:53am UTC](https://discourse.devontechnologies.com/t/item-type-for-references/57943/1 "2020-08-22T03:53:44Z")

</div>

Hi  
I want to save the document type for each file stored in the reference group. There are several ways to do this. One is to create subgroups for each type of resource; For example, books, scientific articles, reports, etc. Another way is to use tags. The third method can be to use custom fields. Given that the use of document type is sometimes used; In my opinion, the use of subgroups prolongs the workflow. But between the second and third methods, I do not know which method is better in the long run and is more compatible with automation scripts.Does anyone have experience in this field. Is there a fourth solution?

---

<div class="post-metadata">

**Author:** ![benoit.pointet](https://discourse.devontechnologies.com/user_avatar/discourse.devontechnologies.com/benoit.pointet/32/9018_2.png) [@benoit.pointet](https://discourse.devontechnologies.com/u/benoit.pointet)\
**Post date:** [August 22, 2020, 6:30am UTC](https://discourse.devontechnologies.com/t/item-type-for-references/57943/2 "2020-08-22T06:30:30Z")

</div>

I use a custom metadata for this.  
Reasoning:

- An item can only be of one document type, so using tags Wouldnt enforce this.
- I display it as a column in list views and use tags for other purposes, thus using tags wouldnt fit.

---

<div class="post-metadata">

**Author:** ![moosavy](https://discourse.devontechnologies.com/letter_avatar_proxy/v4/letter/m/57b2e6/32.png) [@moosavy](https://discourse.devontechnologies.com/u/moosavy)\
**Post date:** [August 22, 2020, 7:03am UTC](https://discourse.devontechnologies.com/t/item-type-for-references/57943/3 "2020-08-22T07:03:09Z")

</div>

with this scenario, can you filter only one document type in DTTG?  
Also with automation scripts, is it handy?

---

<div class="post-metadata">

**Author:** ![benoit.pointet](https://discourse.devontechnologies.com/user_avatar/discourse.devontechnologies.com/benoit.pointet/32/9018_2.png) [@benoit.pointet](https://discourse.devontechnologies.com/u/benoit.pointet)\
**Post date:** [August 22, 2020, 7:05am UTC](https://discourse.devontechnologies.com/t/item-type-for-references/57943/4 "2020-08-22T07:05:49Z")

</div>

Using custom MD is better for scripting, smart groups and smart rules imho. But it does not show up in DTTG.

---

<div class="post-metadata">

**Author:** ![moosavy](https://discourse.devontechnologies.com/letter_avatar_proxy/v4/letter/m/57b2e6/32.png) [@moosavy](https://discourse.devontechnologies.com/u/moosavy)\
**Post date:** [August 22, 2020, 7:07am UTC](https://discourse.devontechnologies.com/t/item-type-for-references/57943/5 "2020-08-22T07:07:08Z")

</div>

Thanks.  
how about embedding to pdf metadata?

---

<div class="post-metadata">

**Author:** ![benoit.pointet](https://discourse.devontechnologies.com/user_avatar/discourse.devontechnologies.com/benoit.pointet/32/9018_2.png) [@benoit.pointet](https://discourse.devontechnologies.com/u/benoit.pointet)\
**Post date:** [August 22, 2020, 7:11am UTC](https://discourse.devontechnologies.com/t/item-type-for-references/57943/6 "2020-08-22T07:11:54Z")

</div>

No clue.

---

<div class="post-metadata">

**Author:** ![BLUEFROG](https://discourse.devontechnologies.com/user_avatar/discourse.devontechnologies.com/bluefrog/32/135_2.png) [@BLUEFROG](https://discourse.devontechnologies.com/u/BLUEFROG)\
**Post date:** [August 22, 2020, 2:21pm UTC](https://discourse.devontechnologies.com/t/item-type-for-references/57943/7 "2020-08-22T14:21:59Z")

</div>

No you can’t filter document types in DEVONthink To Go.

Also, you could add keywords to a PDF _(if you have permissions on the PDF)_ in DEVONthink’s **Document \> Properties** inspector.

---

<div class="post-metadata">

**Author:** ![moosavy](https://discourse.devontechnologies.com/letter_avatar_proxy/v4/letter/m/57b2e6/32.png) [@moosavy](https://discourse.devontechnologies.com/u/moosavy)\
**Post date:** [August 25, 2020, 3:31pm UTC](https://discourse.devontechnologies.com/t/item-type-for-references/57943/8 "2020-08-25T15:31:22Z")

</div>

Thanks.  
How can I add custom fields to pdf properties, like DOI.

---

<div class="post-metadata">

**Author:** ![BLUEFROG](https://discourse.devontechnologies.com/user_avatar/discourse.devontechnologies.com/bluefrog/32/135_2.png) [@BLUEFROG](https://discourse.devontechnologies.com/u/BLUEFROG)\
**Post date:** [August 25, 2020, 3:37pm UTC](https://discourse.devontechnologies.com/t/item-type-for-references/57943/9 "2020-08-25T15:37:05Z")

</div>

Are you running the Pro edition of DEVONthink?

---

<div class="post-metadata">

**Author:** ![moosavy](https://discourse.devontechnologies.com/letter_avatar_proxy/v4/letter/m/57b2e6/32.png) [@moosavy](https://discourse.devontechnologies.com/u/moosavy)\
**Post date:** [August 28, 2020, 5:53am UTC](https://discourse.devontechnologies.com/t/item-type-for-references/57943/10 "2020-08-28T05:53:40Z")

</div>

Yes. Version 3.0.1.

---

<div class="post-metadata">

**Author:** ![cgrunenberg](https://discourse.devontechnologies.com/user_avatar/discourse.devontechnologies.com/cgrunenberg/32/7172_2.png) [@cgrunenberg](https://discourse.devontechnologies.com/u/cgrunenberg)\
**Post date:** [August 28, 2020, 8:01am UTC](https://discourse.devontechnologies.com/t/item-type-for-references/57943/11 "2020-08-28T08:01:59Z")

</div>

The latest version is 3.5.2.

---

<div class="post-metadata">

**Author:** ![moosavy](https://discourse.devontechnologies.com/letter_avatar_proxy/v4/letter/m/57b2e6/32.png) [@moosavy](https://discourse.devontechnologies.com/u/moosavy)\
**Post date:** [August 28, 2020, 8:35am UTC](https://discourse.devontechnologies.com/t/item-type-for-references/57943/12 "2020-08-28T08:35:17Z")

</div>

So, if I upgrade to latest version, changing pdf properties is a native feature?

---

<div class="post-metadata">

**Author:** ![cgrunenberg](https://discourse.devontechnologies.com/user_avatar/discourse.devontechnologies.com/cgrunenberg/32/7172_2.png) [@cgrunenberg](https://discourse.devontechnologies.com/u/cgrunenberg)\
**Post date:** [August 28, 2020, 8:38am UTC](https://discourse.devontechnologies.com/t/item-type-for-references/57943/13 "2020-08-28T08:38:43Z")

</div>

Changing PDF properties is possible since version 2 but requires a visible preview pane (as the document is basically modified).

---

<div class="post-metadata">

**Author:** ![moosavy](https://discourse.devontechnologies.com/letter_avatar_proxy/v4/letter/m/57b2e6/32.png) [@moosavy](https://discourse.devontechnologies.com/u/moosavy)\
**Post date:** [August 31, 2020, 6:41am UTC](https://discourse.devontechnologies.com/t/item-type-for-references/57943/14 "2020-08-31T06:41:55Z")

</div>

I can only add data to available fields. How can I add custom field for example DOI field, to pdf properties?

---

<div class="post-metadata">

**Author:** ![cgrunenberg](https://discourse.devontechnologies.com/user_avatar/discourse.devontechnologies.com/cgrunenberg/32/7172_2.png) [@cgrunenberg](https://discourse.devontechnologies.com/u/cgrunenberg)\
**Post date:** [August 31, 2020, 9:03am UTC](https://discourse.devontechnologies.com/t/item-type-for-references/57943/15 "2020-08-31T09:03:19Z")

</div>

PDF properties can’t be extended but you can use custom metadata (see Info \> Custom inspector and Preferences \> Data). This requires the Pro edition of DEVONthink 3.