Using regexpressions with 'dir' function

47 次查看（过去 30 天）
Galo
Galo 2022-9-21
编辑： dpb 2022-9-21
Hi, I have a folder (MainFolder) that contains other subfolders (Subfolder1, Subfolder2, Subfolder3, ....). Each of those subfolders contain different png images. The format of the name of those images is 'somerandomnameOD1.png' or 'somerandomnameOS1.png'. The last number of the names varies from 1 to 6.
I am trying to use dir function with some regexpressions to extract the path of all those images that end with a pattern like 'OD' or 'OS' followed by the number 1, 2 or 6 (OD1, OD2, OD6, OS1, OS2, OS6) from all the subfolders within the 'MainFolder':
a=dir('C:\folder\folder2\Desktop\MainFolder\**/*O(D|S) (1|2|3).png'));
But the result is an empty struct.
I know I can simply use this:
a=dir('C:\folder\folder2\Desktop\MainFolder\**/*.png'));
and then filter that structure for the names that I need.
But I woudl like to know if it is possible to use reg expressions with a 'dir' function.
Thanks.

请先登录，再回答此问题。

采纳的回答

dpb
dpb 2022-9-21
编辑：dpb 2022-9-21
"I woudl like to know if it is possible to use reg expressions with a 'dir' function."
Nope. Unsupported by the OS; "filename globbing" isn't the same thing as a regular expression.
The '*' and '?' wildcards are supported only.
There are ways by using find or grep but it's simpler to just retrieve with what isolation can be done with the limited wildcards and then apply regexp or other matching tools to that result than to build the commands for the shell.
  1 个评论
Walter Roberson
Walter Roberson 2022-9-21
Right, the suppport for ? and * and ** is provided by dir() itself. dir() is calling into operating system functions to retrieve the contents of the directory, and those functions do not support wildcards or patterns, they just return a list of what is present in the directory, so dir() is handling the filtering.
Historically there has been differences in whether a filename of * means the same thing as *.* on the different operating systems. Unix globbing (filename processing) says that *.* requires that an actual period be present as part of the name, which is different from the historic DOS implementation of 8+3 filenames -- the 8+3 filenames literally do not store the period so 'ABC' and 'ABC.' were indistinguishable because both were stored internally as ABC followed by 5 nulls for the 8 part, and 3 nulls for the extension part. And historically, directories in that time frame had a DIR extension that was routinely hidden, so DEF.DIR was the formal name for folder DEF but eventually the extension for directories stopped being used.
All of which is to say that the processing of wildcards in names is more complicated than one might expect at first, and you need to know which release and which operating system to figure out the precise details. But it has been well over a decade since unix-style filename globbing characters were paid attention to.

请先登录，再进行评论。

更多回答（0 个）

类别

在 Help Center 和 File Exchange 中查找有关 File Operations 的更多信息

标签

Community Treasure Hunt

Find the treasures in MATLAB Central and discover how the community can help you!

Start Hunting!

Translated by