---json
{
    "aliases": [
        { "path": ":howto:tabli:glob" }
    ],
    "description": "With this howto, you will learn what a  glob pattern is and how to use it to select data resources.",
    "low_quality_page": "false",
    "page_id": "s4oqxbqrrgnxp1jvy2kfu"
}
---
====== Tabulify - How to select data resources with a Glob Pattern  ======

===== About =====
With this page, you will learn what a <icon name="carbon:character-patterns"/> [[docs:common:globbing|glob pattern]] is and how to use it to select [[docs:resource:resource|data resources]].


A ''glob pattern'' or ''glob expression'' is a string that can be matched or not against another string. If the ''glob pattern'' matches, a [[docs:resource:resource|data resource]] name, the [[docs:resource:resource|data resource]] is selected otherwise it's not.

To express the pattern, special characters called ''wildcard'' are used that have special meaning. The following paragraphs go through each of this ''wildcard'' and shows you how to use them.

<note>
For the hackers, this is like a regular expression but simplified
</note>




===== Steps =====
==== Prerequisites ====
You should have ''Tabulify'' installed on your computer.

<btn>
[[howto:getting_started:1_install]]
</btn>


==== Star (Any character) ====
The star character ''*'' also known as ''asterix'' matches any number of characters.

If we want to select all [[docs:resource:resource|data resources]] that ends with the term ''sales'', we will use the following ''glob pattern''
<code>
*sales
</code>
where:
  * ''*'' matches all characters before ''sales''
  * ''sales'' matches itself.

Example:
<unit>
<code bash>
tabul data list *sales@tpcds
</code>
<console>
path            media_type
-------------   ------------
catalog_sales   sql/relation
store_sales     sql/relation
web_sales       sql/relation
</console>
</unit>

==== Question mark (One Character) ====
A question mark, ''?'', matches exactly one character.

Example:
  * ''????'' - Matches all data resources with exactly four letters or digits
<unit>
<code bash>
tabul data list ????@tpcds
</code>
<console>
path   media_type
----   ------------
item   sql/relation
</console>
</unit>
  * ''w?*sales'' - Matches any string beginning with ''w'', followed by at least one letter or digit, and ending with ''sales''
<unit>
<code bash>
tabul data list w?*sales@tpcds
</code>
<console>
path        media_type
---------   ------------
web_sales   sql/relation
</console>
</unit>

==== Braces (ie OR) ====

''Braces {}'' specify a collection of subpatterns.

For example:

  * ''{item,store}'' matches ''item'' or ''store''
<unit>
<code bash>
tabul data list {item,store}@tpcds
</code>
<console>
path    media_type
-----   ------------
item    sql/relation
store   sql/relation
</console>
</unit>
  * ''{web*,store*}'' matches all [[docs:resource:resource|data resource]] names that begins with ''web'' or ''store''.
<unit>
<code bash>
tabul data list {web*,store*}@tpcds
</code>
<console>
path            media_type
-------------   ------------
store           sql/relation
store_returns   sql/relation
store_sales     sql/relation
web_page        sql/relation
web_returns     sql/relation
web_sales       sql/relation
web_site        sql/relation
</console>
</unit>

==== Square brackets (Set of characters)====

''Square brackets []'' convey:
  * a set of single characters
  * or a range of characters when the hyphen character (-) is used
that matches a single character.
Within the square brackets, the wildcard ''*'', ''?'', and ''\'' match themselves.

Example:
  * ''[aeiou]'' matches any lowercase vowel.
  * ''[0-9]'' matches any digit.
  * ''[A-Z]'' matches any uppercase letter.
  * ''[a-z,A-Z]'' matches any uppercase or lowercase letter.
  * ''[!abc]'' exclusion does not match the set of characters ''abc''
  * ''*[0-9]*'' ? Matches all strings containing a numeric value

Example:
  * all data resource that  does not have a ''wy'' in their name
<unit>
<code bash>
tabul data list *[wy]*@tpcds
</code>
<console>
path                   media_type
--------------------   ------------
inventory              sql/relation
s_inventory            sql/relation
s_warehouse            sql/relation
s_web_order            sql/relation
s_web_order_lineitem   sql/relation
s_web_page             sql/relation
s_web_returns          sql/relation
s_web_site             sql/relation
warehouse              sql/relation
web_page               sql/relation
web_returns            sql/relation
web_sales              sql/relation
web_site               sql/relation
</console>
</unit>

==== Double Star (Recursive Search) ====

You can also search [[docs:resource:container|container]] such as [[docs:resource:directory|directory]] or [[docs:resource:sql_schema|schema]] recursively by adding to your glob pattern the double start (or asterix).

For instance, the below patter will search the file that starts with ''RE'' in the current directory and all sub-directories of the [[docs:connection:howto|howto connection directory]]
<unit>
<code bash>
tabul data list **/RE*@howto
</code>
  * And there is two ''README.md'' files found that match the pattern
<console>
path                                     media_type
--------------------------------------   ----------
dataset/world/README.md                  text/plain
email/README.md                          text/plain
mysql/coffee-break-sample-db/README.md   text/plain
README.md                                text/plain
recursive/README.md                      text/plain
</console>
</unit>

==== Escape ====
The escape character is the backslash  ''\'' and turns a ''wildcard'' into a simple characters.

Syntax
<code>
\wildcard
</code>

For example:
  * ''\\'' will match a single backslash ''\''
  * ''\?'' matches the question mark ''?''
  * ''\*'' matches the star ''*''

<note>
This is almost always not needed because most of the wildcard are already not allowed when naming resource such as [[docs:resource:file|file]] or [[docs:resource:sql_table|table]].
</note>