better matching, less bugs, more fun

issuecomp-functions.R

@@ -26,20 +26,47 @@ convertLogical0 <- function(var) {
   return(var)
 }
 
-smartPatternMatch <- function(string, pattern, chars) {
-  if(chars < 5) {
-    found <- agrep(pattern, string, max.distance = list(all = 0), ignore.case = TRUE, fixed = FALSE)
+smartPatternMatch <- function(string, pattern, chars, acronym) {
+  pattern <- str_c("\\b", pattern, "\\b")
+  
+  if(chars <= 4) {
+    found <- agrep(pattern, string, max.distance = list(all = 0), ignore.case = !acronym, fixed = FALSE)
   }
-  else if(chars > 7) {
-    found <- agrep(pattern, string, max.distance = list(all = 2), ignore.case = TRUE, fixed = FALSE)
+  else if(chars >= 8) {
+    found <- agrep(pattern, string, max.distance = list(all = 2), ignore.case = !acronym, fixed = FALSE)
   }
   else {
-    found <- agrep(pattern, string, max.distance = list(all = 1), ignore.case = TRUE, fixed = FALSE)
+    found <- agrep(pattern, string, max.distance = list(all = 1), ignore.case = !acronym, fixed = FALSE)
   }
   found <- convertLogical0(found)
   return(found)
 }
 
+viewMatchingTweets <- function(date, issue, folder) {
+  file <- str_c(folder,"/",issue,".csv")
+  df <- read.csv(file, sep = ";", colClasses="character", header = FALSE)
+  for(r in 1:nrow(df)) {
+    curdate <- as.character(df[r,1])
+    if(curdate == date) {
+      curid <- as.character(df[r,2])
+      curtag <- as.character(df[r,3])
+      cat(tweets$text[tweets$id_str == curid]," - ",curtag,"\n")
+    }
+  }
+}
+
+
+checkAcronym <- function(string, chars) {
+  curtag_up <- str_replace_all(string = curtag, pattern = "[[:lower:]]", replacement = "")
+  curchars_up <- nchar(curtag_up, type = "chars")
+  if(curchars_up == curchars) {
+    return(TRUE)
+  }
+  else {
+    return(FALSE)
+  }
+}
+
 ## ERROR HANDLING
 
 # Check for empty API returns (0 or 1 or 2)