Silencing Footnotes in Scientific Books and Papers - RegEx Pattern

Hi, first of all thank you so much for this great app. This is propably the best and most useful app in the whole play store.
 
As I see from the FAQs it is a topic to silence or exchange the uninteresting parts (for listening) of scientific publications. I've tried to understand RegEx fully, but failed. The instruction website is good, but I would need more examples how to implement it in your app, to fully understand the key characters and so on. I hope you can help me with my questions:
 
I have 4 cases which I don't get to work myself:

1. Every sentence starting with a superscript number from 1 to 9999 (with or without space in between) (normaly at the bottom of most pages) should be exchanged
Example: "²Smith, The Book, 1973" or "² Smith, The Book, 1973, p. 12"

2. Every sentence containing "Vgl." or "vgl." should be exchanged
Example: "² Vgl. Smith, 1973" or "²Vgl. Smith, 1973"

3. Superscript Footnotes which are directly behind a "." from 1 to 9999 should be exchanged
Example: "...as shown in his puplication.²"

4. Every Text in "()" containing a year (1900 to 2050) should be exchanged
Example: "(Smith 1973)" or "(Smith, 1973, p. 12)"
Note: This one should be already working with your newest FAQ thread. I'll try that. Thanks.
 
Thanks in advance and all the best
cuibono
 
PS: I have already added many speech edit rules for German and English scientific publications, if there is some interest I can upload my export file here. (rules for stuff like: p. = Page; S. = Seite; etc. = et cetera;...)

Wow thanks for your fast answer. Yes I will try your solution for number 4 in a few days (when I'm back at the office) and will answer here.
 
For the other points: I did use the pdf crop already (which is awesome!), but the publications always have a different number of references on every page (as you suggest) and I wanted to get a better solution for all the publications I want to listen to (often 200-600 pages, that's too much work). In the best case a universal solution of course.
 
I have pdf files, but most of the time use them as a txt in @Voice, because I thought it would be more efficient with your app (cpu usage, errors,...). For txt files the supercript numbers are changed to just normal integer numbers, but I thought if there is a universal solution for superscript in genereal I would try rtf/doc/html or something else (recommondations?). The pdfs on the other side could be different in their handling of superscript, but if there is a universal solution for them it would be the best solution. I'll send you a small collection of examples (just a few pages) in the next days.
 
Thanks a lot in advance!
 
PS: my biggest problem with RegEx is, how to make an if-condition (word) and combine it with a then-condition to remove the whole sentence. There is only one line of code possible. But, I'll send you the examples and I hope from the first working pattern I can understand the rest and every modification needed myself.