Documentation

What's up, Switzerland?

User Tools

Site Tools


02_browsing:04_queries:03_regex

Differences

This shows you the differences between two versions of the page.

Link to this comparison view

Both sides previous revision Previous revision
Next revision
Previous revision
02_browsing:04_queries:03_regex [2020/04/17 20:42]
simone
02_browsing:04_queries:03_regex [2020/05/11 08:56] (current)
Line 1: Line 1:
-====== Regular Expressions ======+====== ​2.4.3 Regular Expressions ======
 In order to search for spelling variants, different forms of a lemma or else, you need to formulate RegEx expressions in ANNIS. For this, you put your query in between slashes. ​ In order to search for spelling variants, different forms of a lemma or else, you need to formulate RegEx expressions in ANNIS. For this, you put your query in between slashes. ​
  
-===== A (very) short introduction to RegEx ===== +In this section we use the following convention:​ 
-"In computing, regular expressions,​ also referred to as RegEx or RegExp, provide a concise and flexible means for matching strings ​of text, such as particular characters, words, or patterns of characters."​ ([[http://en.wikipedia.org/wiki/Regular_expression][Wikipedia]]).+  * Examples for RegEx are in ''​Monospace''​ 
 +  * Whole queries ​as used in ANNIS are in ''/​Monospace with slashes/''​ 
 +  * Results ​of queries are in //italic// 
 +  * Individual letters are in pointy brackets, e.g. <a>
  
-As Wikipedia tells us, RegEx takes a pattern of characters you enter into the search field and looks for matches of these characters in the database. Let us assume that the database to be queried is a string of characters like "the man manually attached the tube in Manchester"​ without any separation of tokens and words. Let us now query the three letters //man//. In this case, RegEx looks for the letter <m> followed by an <a> and then an <n> in the database, regardless of what the pattern is preceded or followed by. As a result, you will get //man// and //manual//, but you will not get //​Manchester//,​ because the RegEx search is case sensitive, see below.+ 
 +===== A (very) short introduction to RegEx ===== 
 +RegEx takes a pattern of characters you enter into the search field and looks for matches of these characters in the database. Let us assume that the database to be queried is a string of characters like "the man manually attached the tube in Manchester"​ without any separation of tokens and words. Let us now query the three letters //man//. In this case, RegEx looks for the letter <m> followed by an <a> and then an <n> in the database, regardless of what the pattern is preceded or followed by. As a result, you will get //man// and //manual//, but you will not get //​Manchester//,​ because the RegEx search is case sensitive, see below.
  
 However, RegEx also allows you to search for such things as alternatives (//man// or //men//), for word boundaries etc. RegEx is a syntax widely spread in programming languages. In what follows, we try to offer an easy overview over the functions you might use most often in this corpus. ​ For more information,​ we refer you to [[http://​www.regular-expressions.info/​regular-expressions.info]]. However, RegEx also allows you to search for such things as alternatives (//man// or //men//), for word boundaries etc. RegEx is a syntax widely spread in programming languages. In what follows, we try to offer an easy overview over the functions you might use most often in this corpus. ​ For more information,​ we refer you to [[http://​www.regular-expressions.info/​regular-expressions.info]].
Line 36: Line 41:
  
 ==Variable letters== ==Variable letters==
-If you are looking for any letter, you can use ''​\w''​ (Remember ​as: word character.)+If you are looking for any letter, you can use ''​\w''​ (remember ​as: word character.)
  
  
Line 55: Line 60:
 E.g ''/​m[A-Z]n/''​ E.g ''/​m[A-Z]n/''​
  
-This search string can also be reduced to e.g. ''​[m-q]''​to find any letter between ​a-q, however useful this may be.+This search string can also be reduced to e.g. ''​[m-q]''​to find any letter between ​<m> and <q>, however useful this may be.
  
-N.B.: <\wcovers all letters from A to z, i.e. uppercase and lowercase. In our corpus, it also includes ​ special letters like äöüàéèß. However, it does not include special characters such as punctuation,​ spaces, <&>​ etc.+N.B.: ''​\w'' ​covers all letters from <Ato <z>, i.e. uppercase and lowercase. In our corpus, it also includes ​ special letters like <äöüàéèß>. However, it does not include special characters such as punctuation,​ spaces, <&>​ etc.
  
  
Line 64: Line 69:
  
 Example: Example:
-m.n+''​m.n''​
 will look for (among others): will look for (among others):
-mAn +  * //mAn// 
-mBn +  * //mBn// 
-man +  * //man// 
-mbn +  * //mbn// 
-m&n +  * //m&n// 
-m n +  * //m n// 
-m_n +  * //m_n// 
-m?n+  * //m?n//
  
  
  
 == Diacritica== == Diacritica==
-This corpus is set up so as to recognize umlauts and letters with accents as individuals (Keep in mind that this is not the case in many other uses of RegEx. Especially in programs that were developed in the US, a <ü> is not considered as a letter but rather as a boundary). ​Searching ​for //mange/will therefore not find any occurrences of //mangé//.+This corpus is set up so as to recognize umlauts and letters with accents as individuals (keep in mind that this is not the case in many other uses of RegEx. Especially in programs that were developed in the US, a <ü> is not considered as a letter but rather as a boundary). ​In our corpus, seearching ​for ''​/mange/'' ​will therefore not find any occurrences of //mangé//.
  
 === Digits=== === Digits===
-Just like <\wabove, you can use <\d> (Remember as: digit) ​to stand in for any digit.+Just like ''​\w'' ​above, you can use ''​\d''  ​to stand in for any digit.
  
 Example: Example:
-n\d+''/​n\d/''​
 will look for (among others): will look for (among others):
-n0 +//n0// 
-n1 +//n1// 
-n9+//n9//
  
  
Line 94: Line 99:
 ==== Separators==== ==== Separators====
 === Individual separating characters=== === Individual separating characters===
-Many different characters can occur in between your letters and digits: ​comas, full stops, spaces etc. Most of these can just be used as they are, i.e:+Many different characters can occur in between your letters and digits: ​commas, full stops, spaces etc. Most of these characters ​can be used for queries like letters or numbers:
    * space    * space
-   ​* ​coma+   ​* ​comma
    * dash (-)    * dash (-)
    * semicolon (;)    * semicolon (;)
Line 105: Line 110:
    * exclamation mark (!)    * exclamation mark (!)
  
-NB: most of these characters do have a special function as well if they appear in a specific position. As you will see below, { } is one of the possible ​way to search for repeating characters. Thus, the character { can be recognized as a character in its own right or as a syntactic function depending on its position. The same goes for most of these characters.+NB: most of these characters do have a special function as well when they appear in a specific position. As you will see below, { } is one of the possible ​ways to search for repeating characters. Thus, the character ​<{can be recognized as a character in its own right or as a syntactic function depending on its position. The same goes for most of these characters.
  
-Other separators are reserved by the RegEx syntax. To use them by their ordinary value, ​you have to escape them, i.e. you have to place a backslash in front of them. Thus, you type in <m\*nto look for //m*n//. These characters are:+Other separators are reserved by the RegEx syntax. To use them by their ordinary value, you have to place a backslash in front of them. Thus, you type in ''/​m\*n/'' ​to look for //m*n//. These characters are:
  
    * asterisk (*)    * asterisk (*)
Line 117: Line 122:
    * dollar ($)    * dollar ($)
    * caret (^)    * caret (^)
-Did we forget anything? Well possible. Just type in the character you are wondering about on its own. If you get an error, you have to escape ​it.+ 
 +In the very probable case this list is not exhaustive, just type in the character you are wondering about. If you get an error, you have to put a backslash in front of it. 
  
 ===Word boundaries=== ===Word boundaries===
-In ANNIS you can query on different layers. ​You cane.g. search for a string of characters in every token or you can search for the same string over whole messages (please keep in mind: this approach is very slow and can result in time-outs!). Depending on which approach you choose for, you have to consider the surrounding environment to your search string.+In ANNIS you can query on different layers. ​For exampleyou can search for a string of characters in every token or you can search for the same string over whole messages (please keep in mind: this approach is very slow and can result in time-outs!). Depending on which approach you choose for, you have to consider the surrounding environment to your search string.
  
-Let us look again at the phrase ​"the man manually attached the tube in Manchester"​. On the **token level**, this phrase ​consists of eight individual tokens: ​+Let us look again at the sentence ​"the man manually attached the tube in Manchester"​. On the **token level**, this sentence ​consists of eight individual tokens: ​
  
 |the|man|manually|attached|the|tube|in|manchester| |the|man|manually|attached|the|tube|in|manchester|
Line 130: Line 137:
 |the man manually attached the tube in Manchester| |the man manually attached the tube in Manchester|
  
-Accordingly,​ the system for querying is different. ​if you query for //man/on the token level, you will find exactly one occurrence, namely the token //man//, because all other tokens contain more than those three characters, e.g. //​manually//​ contains five more characters.+Accordingly,​ the system for querying is different. ​If you query for ''​/man/'' ​on the token level, you will find exactly one occurrence, namely the token //man//, because all other tokens contain more than those three characters, e.g. //​manually//​ contains five more characters.
  
-If you query for //man// on the message level, you will find nothing, because ANNIS will search for a whole message that contains only these three characters. In order to actually find the word you are looking for, you have to query for "any characters followed by the string //man// followed by any characters"​ (the function ​"any characters ​will be introduces later on). Such a string will look like:+If you query for //man// on the message level, you will find nothing, because ANNIS will search for a whole message that contains only these three characters. In order to actually find the word you are looking for, you have to query for "any characters ​(''​.*''​) ​followed by the string //man// followed by any characters"​ (the function ​//any characters// consists of the character full stop that stands for //any character// as shown above. The asterisk stands for an endless repetition as explained in the next section). Such a string will look like:
  
-msg=/.*?man.*/+''​msg=/​.*man.*/​''​
  
 and will find //man// but also //​manually//​. and will find //man// but also //​manually//​.
  
-If you want to find only //man//, you have to query for the three letters surrounded by boundaries (ie. spaces, tabs, fullstops, commas, new-lines etc.). The string for a boundary is //\b//. The query for //man// and only //man// within a message would thus look as follows:+If you want to find only //man//, you have to query for the three letters surrounded by boundaries (ie. spaces, tabs, fullstops, commas, new-lines etc.). The string for a boundary is ''​\b''​. The query for //man// and only //man// within a message would thus look as follows:
  
-msg=/​.*?​\bman\b.*/​+''​msg=/​.*?​\bman\b.*/​''​
  
  
Line 147: Line 154:
  
 ====Quantifiers==== ====Quantifiers====
-Sometimes you might be looking for an expression which can be written with or without repeating letters. E.g. you might want to look for //hallo, haaallo, halooooo// ​etc. in the corpus. Since you do not know how often the individual vowels were repeated, you have to use quantifiers to tell the system that the preceding character can be repeated a certain number of times. Your options are as follows:+Sometimes you might be looking for an expression which can be written with or without repeating letters ​(e.g. you might want to look for //hallo, haaallo, halooooo//). Since you do not know how often the individual vowels were repeated, you have to use quantifiers to tell the system that the preceding character can be repeated a certain number of times. Your options are as follows:
  
-   ​* ​**?** a question mark means a repetition of 1 or 0 times +   ​* ​''​?''​ A question mark means a repetition of 1 or 0 times 
-   * ***** an asterisk means a repetition of 0 or more times +   ​* ​''​*''​ An asterisk means a repetition of 0 or more times 
-   ​* ​**+** an plus sign means a repetition of 1 or more times +   ​* ​''​+''​ A plus sign means a repetition of 1 or more times 
-   ​* ​**{n,m}**  two boundaries in curly brackets mean a repetition of at least n but not more than m times+   ​* ​''​{n,m}'' ​ Two boundaries in curly brackets mean a repetition of at least n but not more than m times
  
  
 Example: Example:
-/​h+a+l+o+/​ +''​/h+a+l+o+/''​ 
-will find all variants of hallo +will find all variants of //hallo//
- +
- +
-Using quantifiers is much more capable and demanding than this. The examples given here are called ​//greedy//, there are also //non greedy quantifiers//​ and //​possessive ones//, which basically look for the last instead of the first occurrence of a token. Please refer to a more explicit manual for these functions. +
- +
-Hint: it you find these options too complicated,​ consider using <​{n,​m}>​ only. With this function, you can fulfill nearly all requirements and it is the easiest function to remember.+
  
  
 ==== Alternatives==== ==== Alternatives====
-Above, you have seen that you can query for different letters in one spot, e.g. you can search for //man// and //men// with the expression ​<m[ae]n>. But what if you want to look for either //n8// or //night// or //nacht// or //nuit//? Here you have to set a <8> equal to the letters <eight><achtand <uit>. To achieve this, you set the whole expression in parentheses and separate the individual variants by a pipe (|).+Above, you have seen that you can query for different letters in one spot, e.g. you can search for //man// and //men// with the expression ​''​m[ae]n''​. But what if you want to look for either //n8// or //night// or //nacht// or //nuit//? Here you have to set a <8> equal to the letter sequence //eight////acht// and //uit//. To achieve this, you set the whole expression in parentheses and separate the individual variants by a pipe (|).
  
 Example: Example:
-n(8|acht|ight|uit) ​+''​(n(8|acht|ight|uit)/'' ​
 will look for: will look for:
-n8 +  * //n8// 
-nacht +  * //nacht// 
-night +  * //night// 
-nuit+  * //nuit//
  
  
  
 ==== A final word==== ==== A final word====
-What you have read hereis only a selection of the possibilities RegEx offers. To keep things more or less simple for you, we tried to document all the features you are likely to use while omitting everything you probably will not care about. Also, there are different implementations of RegEx in different programs and they support different features ​or not. Thus, if you want to use RegEx more intensively or in other places, please read the according manual. If you need more functions, check [[http://​www.regular-expressions.info/​regular-expressions.info]] .+What you have read here is only a selection ​of illustrations ​of the possibilities RegEx offers. To keep things more or less simple for you, we tried to document all the features you are likely to use. Also, there are different implementations of RegEx in different programs and they support different features. Thus, if you want to use RegEx more intensively or in other places, please read the according manual. If you need more functions, ​please ​check  [[http://​www.regular-expressions.info/​regular-expressions.info]].
  
  
02_browsing/04_queries/03_regex.1587148950.txt.gz · Last modified: 2020/05/11 08:55 (external edit)