Miroirs/eliot
1
0
Fork 0
mirror of git://git.savannah.nongnu.org/eliot.git synced 2025-01-17 06:11:49 +01:00
Code Issues Projects Releases Packages Wiki Activity

some more doc on dictionary and expected regexp syntax

Browse source
This commit is contained in:
Antoine Fraboulet 2005-04-19 16:25:06 +00:00
parent 1077565ace
commit b1e67d947b
2 changed files with 22 additions and 8 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

16
doc/dic.txt
View file

@ -73,7 +73,7 @@ The header structure is the following:
#define _COMPIL_KEYWORD_ "_COMPILED_DICTIONARY_"
typedef struct _Dict_header {
typedef struct _Dict_header { // offset
char ident[sizeof(_COMPIL_KEYWORD_)]; // 0x00
char unused_1; // 0x16
char unused_2; // 0x17
@ -82,7 +82,7 @@ typedef struct _Dict_header {
unsigned int edgesused; // 0x20
unsigned int nodesused; // 0x24
unsigned int nodessaved; // 0x2c
unsigned int edgessaved; // 0x28
unsigned int edgessaved; // 0x30
} Dict_header;
binary output of the header:
@ -98,11 +98,11 @@ binary output of the header:
0x2c edges saved : 1 00000001
===================================================================
The real array of data begins at offset 0x30. Integer are stored in a
The real array of data begins at offset 0x34. Integer are stored in a
machine dependent way. This dictionary was compiled on a i386 and is
not readable on a machine with a different endianess. The array is
stored 'as is' right after the header. Each array cell is a
bit-structure:
not readable on a machine with a different endianess (unless swapping
all necessary information). The array is stored 'as is' right after
the header. Each array cell is a bit-structure on 4 bytes :
typedef struct _Dawg_edge {
unsigned int ptr : 24;
@ -115,8 +115,8 @@ typedef struct _Dawg_edge {
Characters are not stored in ASCII. The order is preserved but
we changed the values: A=1, B=2, ... This is very easy to do
with the ASCII table as ('A' & 0x1f) == ('a' & 0x1f) == 1.
This may not work on machines that are not using ASCII (like
Macintosh.)
This may not work on machines that are not using ASCII. The dictionary
can thus handle up to 32 different letters but not more.
offs binary structure
---- -------- | ------------------

14
doc/regexp.tex
View file

@ -8,6 +8,10 @@ expressions rationnelles habituelles.
\section{utilisation}
Les mots recherchés sont complets : la recherche d'une expression
\verb=e= correspond à l'expression \verb=^e$=. Pour rechercher un
motif \verb=m= dans un mot il faut donc utiliser l'expression \verb=.*m.*=
\subsection{caractères}
\begin{itemize}
@ -15,6 +19,8 @@ expressions rationnelles habituelles.
\item \texttt{.} : n'importe quel caractère
\item \texttt{:v:} : n'importe quelle voyelle
\item \texttt{:c:} : n'importe quelle consonne
\item \texttt{:1:} : liste 1 définie par l'utilisateur
\item \texttt{:2:} : liste 2 définie par l'utilisateur
\end{itemize}
\subsection{répétitions}
@ -38,4 +44,12 @@ expressions rationnelles habituelles.
\subsection{exemples}
\begin{itemize}
\item \verb=a.*= : liste des mots débutant par la lettre \verb=a=
\item \verb=.*a= : liste des mots se terminant par la lettre \verb=a=
\item \verb=.*oula.*= : liste des mots contenant le motif \verb=oula=
\item \verb=a.*b= : liste des mots débutant par \verb=a= et se terminant par \verb=b=
\item \verb=.*a.*e.*i.*o.*u.*= : liste des mots contenant les les lettres \verb=aeiou= dans l'ordre
\end{itemize}
\end{document}