SCM

SCM Repository

[tm] Diff of /pkg/R/corpus.R
ViewVC logotype

Diff of /pkg/R/corpus.R

Parent Directory Parent Directory | Revision Log Revision Log | View Patch Patch

trunk/R/textmin/R/textdoccol.R revision 61, Mon Oct 23 20:07:05 2006 UTC pkg/R/corpus.R revision 1114, Fri Nov 26 14:05:54 2010 UTC
# Line 1  Line 1 
1  # Author: Ingo Feinerer  # Author: Ingo Feinerer
2    
3  setGeneric("TextDocCol", function(object, parser = plaintext.parser, lod = FALSE) standardGeneric("TextDocCol"))  .PCorpus <- function(x, cmeta, dmeta, dbcontrol) {
4  setMethod("TextDocCol",      attr(x, "CMetaData") <- cmeta
5            signature(object = "character"),      attr(x, "DMetaData") <- dmeta
6            function(object, parser = plaintext.parser, lod = FALSE) {      attr(x, "DBControl") <- dbcontrol
7                filelist <- dir(object, full.names = TRUE)      class(x) <- c("PCorpus", "Corpus", "list")
8                tdl <- lapply(filelist, parser, lod)      x
9                return(new("TextDocCol", .Data = tdl))  }
10            })  DBControl <- function(x) attr(x, "DBControl")
11    
12  plaintext.parser <- function(file, lod) {  PCorpus <- function(x,
13      id <- file                      readerControl = list(reader = x$DefaultReader, language = "en"),
14      origin <- dirname(file)                      dbControl = list(dbName = "", dbType = "DB1"),
15                        ...) {
16        readerControl <- prepareReader(readerControl, x$DefaultReader, ...)
17    
18        if (is.function(readerControl$init))
19            readerControl$init()
20    
21        if (is.function(readerControl$exit))
22            on.exit(readerControl$exit())
23    
24        if (!filehash::dbCreate(dbControl$dbName, dbControl$dbType))
25            stop("error in creating database")
26        db <- filehash::dbInit(dbControl$dbName, dbControl$dbType)
27    
28        # Allocate memory in advance if length is known
29        tdl <- if (x$Length > 0)
30            vector("list", as.integer(x$Length))
31        else
32            list()
33    
34      doc <- new("PlainTextDocument", FileName = file, Cached = FALSE, Author = "Unknown",      counter <- 1
35                 DateTimeStamp = date(), Description = "", ID = id, Origin = origin, Heading = "")      while (!eoi(x)) {
36            x <- stepNext(x)
37            elem <- getElem(x)
38            doc <- readerControl$reader(elem, readerControl$language, if (is.null(x$Names)) as.character(counter) else x$Names[counter])
39            filehash::dbInsert(db, ID(doc), doc)
40            if (x$Length > 0) tdl[[counter]] <- ID(doc)
41            else tdl <- c(tdl, ID(doc))
42            counter <- counter + 1
43        }
44        names(tdl) <- x$Names
45    
46        df <- data.frame(MetaID = rep(0, length(tdl)), stringsAsFactors = FALSE)
47        filehash::dbInsert(db, "DMetaData", df)
48        dmeta.df <- data.frame(key = "DMetaData", subset = I(list(NA)))
49    
50        .PCorpus(tdl, .MetaDataNode(), dmeta.df, dbControl)
51    }
52    
53    .VCorpus <- function(x, cmeta, dmeta) {
54        attr(x, "CMetaData") <- cmeta
55        attr(x, "DMetaData") <- dmeta
56        class(x) <- c("VCorpus", "Corpus", "list")
57        x
58    }
59    
60    # Register S3 corpus classes to be recognized by S4 methods. This is
61    # mainly a fix to be compatible with packages which were originally
62    # developed to cooperate with corresponding S4 tm classes. Necessary
63    # since tm's class architecture was changed to S3 since tm version 0.5.
64    setOldClass(c("VCorpus", "Corpus", "list"))
65    
66    # The "..." are additional arguments for the FunctionGenerator reader
67    VCorpus <- Corpus <- function(x,
68                                  readerControl = list(reader = x$DefaultReader, language = "en"),
69                                  ...) {
70        readerControl <- prepareReader(readerControl, x$DefaultReader, ...)
71    
72        if (is.function(readerControl$init))
73            readerControl$init()
74    
75        if (is.function(readerControl$exit))
76            on.exit(readerControl$exit())
77    
78        # Allocate memory in advance if length is known
79        tdl <- if (x$Length > 0)
80            vector("list", as.integer(x$Length))
81        else
82            list()
83    
84      if (lod) {      if (x$Vectorized)
85          doc <- loadFileIntoMem(doc)          tdl <- mapply(function(x, id) readerControl$reader(x, readerControl$language, id),
86                          pGetElem(x),
87                          id = if (is.null(x$Names)) as.character(seq_len(x$Length)) else x$Names,
88                          SIMPLIFY = FALSE)
89        else {
90            counter <- 1
91            while (!eoi(x)) {
92                x <- stepNext(x)
93                elem <- getElem(x)
94                doc <- readerControl$reader(elem, readerControl$language, if (is.null(x$Names)) as.character(counter) else x$Names[counter])
95                if (x$Length > 0)
96                    tdl[[counter]] <- doc
97                else
98                    tdl <- c(tdl, list(doc))
99                counter <- counter + 1
100            }
101        }
102        names(tdl) <- x$Names
103        df <- data.frame(MetaID = rep(0, length(tdl)), stringsAsFactors = FALSE)
104        .VCorpus(tdl, .MetaDataNode(), df)
105      }      }
106    
107      return(doc)  `[.PCorpus` <- function(x, i) {
108        if (missing(i)) return(x)
109        index <- attr(x, "DMetaData")[[1 , "subset"]]
110        attr(x, "DMetaData")[[1 , "subset"]] <- if (is.numeric(index)) index[i] else i
111        dmeta <- attr(x, "DMetaData")
112        .PCorpus(NextMethod("["), CMetaData(x), dmeta, DBControl(x))
113  }  }
114    
115  reuters21578xml.parser <- function(file, lod) {  `[.VCorpus` <- function(x, i) {
116      tree <- xmlTreeParse(file)      if (missing(i)) return(x)
117      node <- xmlRoot(tree)      .VCorpus(NextMethod("["), CMetaData(x), DMetaData(x)[i, , drop = FALSE])
118    }
119    
120      # The <AUTHOR></AUTHOR> tag is unfortunately NOT obligatory!  `[<-.PCorpus` <- function(x, i, value) {
121      if (!is.null(node[["TEXT"]][["AUTHOR"]]))      db <- filehash::dbInit(DBControl(x)[["dbName"]], DBControl(x)[["dbType"]])
122          author <- xmlValue(node[["TEXT"]][["AUTHOR"]])      counter <- 1
123      else      for (id in unclass(x)[i]) {
124          author <- ""          if (identical(length(value), 1L)) db[[id]] <- value
125            else db[[id]] <- value[[counter]]
126            counter <- counter + 1
127        }
128        x
129    }
130    
131      datetimestamp <- xmlValue(node[["DATE"]])  .map_name_index <- function(x, i) {
132      description <- ""      if (is.character(i)) {
133      id <- xmlAttrs(node)[["NEWID"]]          if (is.null(names(x)))
134                match(i, meta(x, "ID", type = "local"))
     # The <TITLE></TITLE> tag is unfortunately NOT obligatory!  
     if (!is.null(node[["TEXT"]][["TITLE"]]))  
         heading <- xmlValue(node[["TEXT"]][["TITLE"]])  
135      else      else
136          heading <- ""              match(i, names(x))
   
     topics <- unlist(xmlApply(node[["TOPICS"]], function(x) xmlValue(x)), use.names = FALSE)  
   
     doc <- new("XMLTextDocument", FileName = file, Cached = FALSE, Author = author,  
                DateTimeStamp = datetimestamp, Description = "", ID = id, Origin = "Reuters-21578 XML",  
                Heading = heading, LocalMetaData = list(Topics = topics))  
   
     if (lod) {  
         doc <- loadFileIntoMem(doc)  
137      }      }
138        i
     return(doc)  
139  }  }
140    
141  rcv1.parser <- function(file, lod) {  `[[.PCorpus` <-  function(x, i) {
142      tree <- xmlTreeParse(file)      i <- .map_name_index(x, i)
143      node <- xmlRoot(tree)      db <- filehash::dbInit(DBControl(x)[["dbName"]], DBControl(x)[["dbType"]])
144        filehash::dbFetch(db, NextMethod("[["))
145    }
146    `[[.VCorpus` <-  function(x, i) {
147        i <- .map_name_index(x, i)
148        lazyTmMap <- meta(x, tag = "lazyTmMap", type = "corpus")
149        if (!is.null(lazyTmMap))
150            .Call("copyCorpus", x, materialize(x, i))
151        NextMethod("[[")
152    }
153    
154      datetimestamp <- xmlAttrs(node)[["date"]]  `[[<-.PCorpus` <-  function(x, i, value) {
155      id <- xmlAttrs(node)[["itemid"]]      i <- .map_name_index(x, i)
156      heading <- xmlValue(node[["title"]])      db <- filehash::dbInit(DBControl(x)[["dbName"]], DBControl(x)[["dbType"]])
157        index <- unclass(x)[[i]]
158        db[[index]] <- value
159        x
160    }
161    `[[<-.VCorpus` <-  function(x, i, value) {
162        i <- .map_name_index(x, i)
163        # Mark new objects as not active for lazy mapping
164        lazyTmMap <- meta(x, tag = "lazyTmMap", type = "corpus")
165        if (!is.null(lazyTmMap)) {
166            lazyTmMap$index[i] <- FALSE
167            meta(x, tag = "lazyTmMap", type = "corpus") <- lazyTmMap
168        }
169        # Set the value
170        cl <- class(x)
171        y <- NextMethod("[[<-")
172        class(y) <- cl
173        y
174    }
175    
176      doc <- new("XMLTextDocument", FileName = file, Cached = FALSE, Author = "",  # Update NodeIDs of a CMetaData tree
177                 DateTimeStamp = datetimestamp, Description = "", ID = id, Origin = "Reuters Corpus Volume 1 XML",  .update_id <- function(x, id = 0, mapping = NULL, left.mapping = NULL, level = 0) {
178                 Heading = heading)      # Traversal of (binary) CMetaData tree with setup of NodeIDs
179        set_id <- function(x) {
180            x$NodeID <- id
181            id <<- id + 1
182            level <<- level + 1
183            if (length(x$Children) > 0) {
184                mapping <<- cbind(mapping, c(x$Children[[1]]$NodeID, id))
185                left <- set_id(x$Children[[1]])
186                if (level == 1) {
187                    left.mapping <<- mapping
188                    mapping <<- NULL
189                }
190                mapping <<- cbind(mapping, c(x$Children[[2]]$NodeID, id))
191                right <- set_id(x$Children[[2]])
192    
193      if (lod) {              x$Children <- list(left, right)
194          doc <- loadFileIntoMem(doc)          }
195            level <<- level - 1
196            x
197        }
198        list(root = set_id(x), left.mapping = left.mapping, right.mapping = mapping)
199      }      }
200    
201      return(doc)  # Find indices to be updated for a CMetaData tree
202    .find_indices <- function(x) {
203        indices.mapping <- NULL
204        for (m in levels(as.factor(DMetaData(x)$MetaID))) {
205            indices <- (DMetaData(x)$MetaID == m)
206            indices.mapping <- c(indices.mapping, list(m = indices))
207            names(indices.mapping)[length(indices.mapping)] <- m
208        }
209        indices.mapping
210  }  }
211    
212  uci.kdd.newsgroup.parser <-  function(file, lod) {  c2 <- function(x, y, ...) {
213      mail <- readLines(file)      # Update the CMetaData tree
214      author <- gsub("From: ", "", grep("^From:", mail, value = TRUE))      cmeta <- .MetaDataNode(0, list(merge_date = as.POSIXlt(Sys.time(), tz = "GMT"), merger = Sys.getenv("LOGNAME")), list(CMetaData(x), CMetaData(y)))
215      datetimestamp <- gsub("Date: ", "", grep("^Date:", mail, value = TRUE))      update.struct <- .update_id(cmeta)
     origin <- gsub("Path: ", "", grep("^Path:", mail, value = TRUE))  
     heading <- gsub("Subject: ", "", grep("^Subject:", mail, value = TRUE))  
     newsgroup <- gsub("Newsgroups: ", "", grep("^Newsgroups:", mail, value = TRUE))  
216    
217      new("NewsgroupDocument", FileName = file, Cached = FALSE, Author = author, DateTimeStamp = datetimestamp,      new <- .VCorpus(c(unclass(x), unclass(y)), update.struct$root, NULL)
         Description = "", ID = file, Origin = origin, Heading = heading, Newsgroup = newsgroup)  
218    
219      if (lod) {      # Find indices to be updated for the left tree
220          doc <- loadFileIntoMem(doc)      indices.mapping <- .find_indices(x)
     }  
221    
222      return(doc)      # Update the DMetaData data frames for the left tree
223        for (i in 1:ncol(update.struct$left.mapping)) {
224            map <- update.struct$left.mapping[,i]
225            DMetaData(x)$MetaID <- replace(DMetaData(x)$MetaID, indices.mapping[[as.character(map[1])]], map[2])
226  }  }
227    
228  # Parse a <newsitem></newsitem> element from a well-formed RCV1 XML file      # Find indices to be updated for the right tree
229  rcv1.to.plain <- function(node) {      indices.mapping <- .find_indices(y)
     datetimestamp <- xmlAttrs(node)[["date"]]  
     id <- xmlAttrs(node)[["itemid"]]  
     origin <- "Reuters Corpus Volume 1 XML"  
     corpus <- unlist(xmlApply(node[["text"]], xmlValue), use.names = FALSE)  
     heading <- xmlValue(node[["title"]])  
230    
231      new("PlainTextDocument", .Data = corpus, Author = "", DateTimeStamp = datetimestamp,      # Update the DMetaData data frames for the right tree
232          Description = "", ID = id, Origin = "Reuters Corpus Volume 1 XML", Heading = heading)      for (i in 1:ncol(update.struct$right.mapping)) {
233            map <- update.struct$right.mapping[,i]
234            DMetaData(y)$MetaID <- replace(DMetaData(y)$MetaID, indices.mapping[[as.character(map[1])]], map[2])
235  }  }
236    
237  # Parse a <REUTERS></REUTERS> element from a well-formed Reuters-21578 XML file      # Merge the DMetaData data frames
238  reuters21578xml.to.plain <- function(node) {      labels <- setdiff(names(DMetaData(y)), names(DMetaData(x)))
239      # The <AUTHOR></AUTHOR> tag is unfortunately NOT obligatory!      na.matrix <- matrix(NA, nrow = nrow(DMetaData(x)), ncol = length(labels), dimnames = list(row.names(DMetaData(x)), labels))
240      if (!is.null(node[["TEXT"]][["AUTHOR"]]))      x.dmeta.aug <- cbind(DMetaData(x), na.matrix)
241          author <- xmlValue(node[["TEXT"]][["AUTHOR"]])      labels <- setdiff(names(DMetaData(x)), names(DMetaData(y)))
242      else      na.matrix <- matrix(NA, nrow = nrow(DMetaData(y)), ncol = length(labels), dimnames = list(row.names(DMetaData(y)), labels))
243          author <- ""      y.dmeta.aug <- cbind(DMetaData(y), na.matrix)
244        DMetaData(new) <- rbind(x.dmeta.aug, y.dmeta.aug)
245    
246      datetimestamp <- xmlValue(node[["DATE"]])      new
247      description <- ""  }
     id <- xmlAttrs(node)[["NEWID"]]  
248    
249      origin <- "Reuters-21578 XML"  c.Corpus <-
250    function(x, ..., recursive = FALSE)
251    {
252        args <- list(...)
253    
254      # The <BODY></BODY> tag is unfortunately NOT obligatory!      if (identical(length(args), 0L))
255      if (!is.null(node[["TEXT"]][["BODY"]]))          return(x)
         corpus <- xmlValue(node[["TEXT"]][["BODY"]])  
     else  
         corpus <- ""  
256    
257      # The <TITLE></TITLE> tag is unfortunately NOT obligatory!      if (!all(unlist(lapply(args, inherits, class(x)))))
258      if (!is.null(node[["TEXT"]][["TITLE"]]))          stop("not all arguments are of the same corpus type")
         heading <- xmlValue(node[["TEXT"]][["TITLE"]])  
     else  
         heading <- ""  
259    
260      topics <- unlist(xmlApply(node[["TOPICS"]], function(x) xmlValue(x)), use.names = FALSE)      if (inherits(x, "PCorpus"))
261            stop("concatenation of corpora with underlying databases is not supported")
262    
263      new("PlainTextDocument", .Data = corpus, Cached = TRUE, Author = author, DateTimeStamp = datetimestamp,      l <- base::c(list(x), args)
264          Description = description, ID = id, Origin = origin, Heading = heading, LocalMetaData = list(Topics = topics))      if (recursive)
265            Reduce(c2, l)
266        else {
267            l <- do.call("c", lapply(l, unclass))
268            .VCorpus(l,
269                     cmeta = .MetaDataNode(),
270                     dmeta = data.frame(MetaID = rep(0, length(l)), stringsAsFactors = FALSE))
271        }
272  }  }
273    
274  setGeneric("loadFileIntoMem", function(object) standardGeneric("loadFileIntoMem"))  c.TextDocument <- function(x, ..., recursive = FALSE) {
275  setMethod("loadFileIntoMem",      args <- list(...)
           c("PlainTextDocument"),  
           function(object) {  
               if (!Cached(object)) {  
                   corpus <- readLines(FileName(object))  
                   Corpus(object) <- corpus  
                   Cached(object) <- TRUE  
                   return(object)  
               } else {  
                   return(object)  
               }  
           })  
 setMethod("loadFileIntoMem",  
           c("XMLTextDocument"),  
           function(object) {  
               if (!Cached(object)) {  
                   file <- FileName(object)  
                   doc <- xmlTreeParse(file)  
                   class(doc) <- "list"  
                   Corpus(object) <- doc  
                   Cached(object) <- TRUE  
                   return(object)  
               } else {  
                   return(object)  
               }  
           })  
 setMethod("loadFileIntoMem",  
           c("NewsgroupDocument"),  
           function(object) {  
               if (!Cached(object)) {  
                   mail <- readLines(FileName(object))  
                   Cached(object) <- TRUE  
                   index <- grep("^Lines:", mail)  
                   Corpus(object) <- mail[(index + 1):length(mail)]  
                   return(object)  
               } else {  
                   return(object)  
               }  
           })  
   
 setGeneric("tm_transform", function(object, FUN, ...) standardGeneric("tm_transform"))  
 setMethod("tm_transform",  
           c("TextDocCol"),  
           function(object, FUN, ...) {  
               result <- as(lapply(object, FUN, ..., GlobalMetaData = GlobalMetaData(object)), "TextDocCol")  
               result@GlobalMetaData <- GlobalMetaData(object)  
               return(result)  
           })  
   
 setGeneric("toPlainTextDocument", function(object, FUN, ...) standardGeneric("toPlainTextDocument"))  
 setMethod("toPlainTextDocument",  
           c("PlainTextDocument"),  
           function(object, FUN, ...) {  
               return(object)  
           })  
 setMethod("toPlainTextDocument",  
           c("XMLTextDocument"),  
           function(object, FUN, ...) {  
               if (!Cached(object))  
                   object <- loadFileIntoMem(object)  
   
               corpus <- Corpus(object)  
   
               # As XMLDocument is no native S4 class, restore valid information  
               class(corpus) <- "XMLDocument"  
               names(corpus) <- c("doc","dtd")  
   
               return(FUN(xmlRoot(corpus), ...))  
           })  
   
 setGeneric("stemTextDocument", function(object, ...) standardGeneric("stemTextDocument"))  
 setMethod("stemTextDocument",  
           c("PlainTextDocument"),  
           function(object, ...) {  
               if (!Cached(object))  
                   object <- loadFileIntoMem(object)  
   
               require(Rstem)  
               splittedCorpus <- unlist(strsplit(object, " ", fixed = TRUE))  
               stemmedCorpus <- wordStem(splittedCorpus, ...)  
               Corpus(object) <- paste(stemmedCorpus, collapse = " ")  
               return(object)  
           })  
   
 setGeneric("removeStopWords", function(object, stopwords, ...) standardGeneric("removeStopWords"))  
 setMethod("removeStopWords",  
           signature(object = "PlainTextDocument", stopwords = "character"),  
           function(object, stopwords, ...) {  
               if (!Cached(object))  
                   object <- loadFileIntoMem(object)  
   
               require(Rstem)  
               splittedCorpus <- unlist(strsplit(object, " ", fixed = TRUE))  
               noStopwordsCorpus <- splittedCorpus[!splittedCorpus %in% stopwords]  
               Corpus(object) <- paste(noStopwordsCorpus, collapse = " ")  
               return(object)  
           })  
   
 setGeneric("tm_filter", function(object, ..., FUN = s.filter) standardGeneric("tm_filter"))  
 setMethod("tm_filter",  
           signature(object = "TextDocCol"),  
           function(object, ..., FUN = s.filter) {  
               object[tm_index(object, ..., FUN)]  
           })  
   
 setGeneric("tm_index", function(object, ..., FUN = s.filter) standardGeneric("tm_index"))  
 setMethod("tm_index",  
           signature(object = "TextDocCol"),  
           function(object, ..., FUN = s.filter) {  
               sapply(object, FUN, ..., GlobalMetaData = GlobalMetaData(object))  
           })  
   
 s.filter <- function(object, s, ..., GlobalMetaData) {  
     b <- TRUE  
     for (tag in names(s)) {  
         if (tag %in% names(LocalMetaData(object))) {  
             b <- b && any(grep(s[[tag]], LocalMetaData(object)[[tag]]))  
         } else if (tag %in% names(GlobalMetaData)){  
             b <- b && any(grep(s[[tag]], GlobalMetaData[[tag]]))  
         } else {  
             b <- b && any(grep(s[[tag]], eval(call(tag, object))))  
         }  
     }  
     return(b)  
 }  
   
 setGeneric("fulltext.search.filter", function(object, pattern, ...) standardGeneric("fulltext.search.filter"))  
 setMethod("fulltext.search.filter",  
           signature(object = "PlainTextDocument", pattern = "character"),  
           function(object, pattern, ...) {  
               if (!Cached(object))  
                   object <- loadFileIntoMem(object)  
   
               return(any(grep(pattern, Corpus(object))))  
           })  
   
 setGeneric("reuters21578.topic.filter", function(object, topics, ...) standardGeneric("reuters21578.topic.filter"))  
 setMethod("reuters21578.topic.filter",  
           c("PlainTextDocument", "character"),  
           function(object, topics, ...) {  
               if (!Cached(object))  
                   object <- loadFileIntoMem(object)  
276    
277                if (any(LocalMetaData(object)$Topics %in% topics))      if (identical(length(args), 0L))
278                    return(TRUE)          return(x)
               else  
                   return(FALSE)  
           })  
279    
280  setGeneric("id.filter", function(object, IDs, ...) standardGeneric("id.filter"))      if (!all(unlist(lapply(args, inherits, class(x)))))
281  setMethod("id.filter",          stop("not all arguments are text documents")
           c("TextDocument", "numeric"),  
           function(object, IDs, ...) {  
               if (ID(object) %in% IDs)  
                   return(TRUE)  
               else  
                   return(FALSE)  
           })  
282    
283  setGeneric("attachData", function(object, data) standardGeneric("attachData"))      dmeta <- data.frame(MetaID = rep(0, length(list(x, ...))), stringsAsFactors = FALSE)
284  setMethod("attachData",      .VCorpus(list(x, ...), .MetaDataNode(), dmeta)
285            c("TextDocCol","TextDocument"),  }
           function(object, data) {  
               data <- as(list(data), "TextDocCol")  
               object@.Data <- as(c(object@.Data, data), "TextDocCol")  
               return(object)  
           })  
   
 setGeneric("attachMetaData", function(object, name, metadata) standardGeneric("attachMetaData"))  
 setMethod("attachMetaData",  
           c("TextDocCol"),  
           function(object, name, metadata) {  
               object@GlobalMetaData <- c(GlobalMetaData(object), new = list(metadata))  
               names(object@GlobalMetaData)[length(names(GlobalMetaData(object)))] <- name  
               return(object)  
           })  
   
 setGeneric("setSubscriptable", function(object, name) standardGeneric("setSubscriptable"))  
 setMethod("setSubscriptable",  
           c("TextDocCol"),  
           function(object, name) {  
               if (!is.character(GlobalMetaData(object)$subscriptable))  
                   object <- attachMetaData(object, "subscriptable", name)  
               else  
                   object@GlobalMetaData$subscriptable <- c(GlobalMetaData(object)$subscriptable, name)  
               return(object)  
           })  
   
 setMethod("[",  
           signature(x = "TextDocCol", i = "ANY", j = "ANY", drop = "ANY"),  
           function(x, i, j, ... , drop) {  
               if(missing(i))  
                   return(x)  
286    
287                object <- x  print.Corpus <- function(x, ...) {
288                object@.Data <- x@.Data[i, ..., drop = FALSE]      cat(sprintf(ngettext(length(x),
289                for (m in names(GlobalMetaData(object))) {                           "A corpus with %d text document\n",
290                    if (m %in% GlobalMetaData(object)$subscriptable) {                           "A corpus with %d text documents\n"),
291                        object@GlobalMetaData[[m]] <- GlobalMetaData(object)[[m]][i, ..., drop = FALSE]                  length(x)))
292        invisible(x)
293    }
294    
295    summary.Corpus <- function(object, ...) {
296        print(object)
297        if (length(DMetaData(object)) > 0) {
298            cat(sprintf(ngettext(length(attr(CMetaData(object), "MetaData")),
299                                 "\nThe metadata consists of %d tag-value pair and a data frame\n",
300                                 "\nThe metadata consists of %d tag-value pairs and a data frame\n"),
301                        length(CMetaData(object)$MetaData)))
302            cat("Available tags are:\n")
303            cat(strwrap(paste(names(CMetaData(object)$MetaData), collapse = " "), indent = 2, exdent = 2), "\n")
304            cat("Available variables in the data frame are:\n")
305            cat(strwrap(paste(names(DMetaData(object)), collapse = " "), indent = 2, exdent = 2), "\n")
306                    }                    }
307                }                }
               return(object)  
           })  
   
 setMethod("c",  
           signature(x = "TextDocCol"),  
           function(x, ..., recursive = TRUE){  
               args <- list(...)  
               if(length(args) == 0)  
                   return(x)  
               return(as(c(as(x, "list"), ...), "TextDocCol"))  
     })  
308    
309  setMethod("length",  inspect <- function(x) UseMethod("inspect", x)
310            signature(x = "TextDocCol"),  inspect.PCorpus <- function(x) {
311            function(x){      summary(x)
               return(length(as(x, "list")))  
     })  
   
 setMethod("show",  
           signature(object = "TextDocCol"),  
           function(object){  
               cat("A text document collection with", length(object), "text document")  
               if (length(object) == 1)  
312                    cat("\n")                    cat("\n")
313                else      db <- filehash::dbInit(DBControl(x)[["dbName"]], DBControl(x)[["dbType"]])
314                    cat("s\n")      show(filehash::dbMultiFetch(db, unlist(x)))
     })  
   
 setMethod("summary",  
           signature(object = "TextDocCol"),  
           function(object){  
               show(object)  
               if (length(GlobalMetaData(object)) > 0) {  
                   cat("\nThe global metadata consists of", length(GlobalMetaData(object)), "tag-value pair")  
                   if (length(GlobalMetaData(object)) == 1)  
                       cat(".\n")  
                   else  
                       cat("s.\n")  
                   cat("Available tags are:\n")  
                   cat(names(GlobalMetaData(object)), "\n")  
315                }                }
316      })  inspect.VCorpus <- function(x) {
317        summary(x)
 setGeneric("inspect", function(object) standardGeneric("inspect"))  
 setMethod("inspect",  
           c("TextDocCol"),  
           function(object) {  
               summary(object)  
318                cat("\n")                cat("\n")
319                show(as(object, "list"))      print(noquote(lapply(x, identity)))
320            })  }
321    
322    lapply.PCorpus <- function(X, FUN, ...) {
323        db <- filehash::dbInit(DBControl(X)[["dbName"]], DBControl(X)[["dbType"]])
324        lapply(filehash::dbMultiFetch(db, unlist(X)), FUN, ...)
325    }
326    lapply.VCorpus <- function(X, FUN, ...) {
327        lazyTmMap <- meta(X, tag = "lazyTmMap", type = "corpus")
328        if (!is.null(lazyTmMap))
329            .Call("copyCorpus", X, materialize(X))
330        base::lapply(X, FUN, ...)
331    }
332    
333    writeCorpus <-  function(x, path = ".", filenames = NULL) {
334        filenames <- file.path(path,
335                               if (is.null(filenames)) unlist(lapply(x, function(x) sprintf("%s.txt", ID(x))))
336                               else filenames)
337        i <- 1
338        for (o in x) {
339            writeLines(as.PlainTextDocument(o), filenames[i])
340            i <- i + 1
341        }
342    }

Legend:
Removed from v.61  
changed lines
  Added in v.1114

root@r-forge.r-project.org
ViewVC Help
Powered by ViewVC 1.0.0  
Thanks to:
Vienna University of Economics and Business Powered By FusionForge