P$get_text_from_boxes(paragraphs = FALSE) in seemingly normal document

[ChristophLeonhardt]

After parsing 142 optically identical documents, get_text_from_boxes throws following error:

Fehler in if (position["top"] > box["top"] && position["top"] < box["bottom"] && :
Fehlender Wert, wo TRUE/FALSE nötig ist

Document can be found online here: https://www.landtag-saar.de/Plenarprotokoll/PlPr15_041.pdf

Used box coordinates:

  P$add_box(box = c(top = 105, height = 690, left = 45, width = 250), page = NULL, replace = TRUE)
  P$add_box(box = c(top = 105, height = 690, left = 297, width = 250), page = NULL, replace = FALSE)
  
  P$add_box(box = c(top = 86, height = 700, left = 45, width = 250), page = 1, replace = TRUE)
  P$add_box(box = c(top = 86, height = 700, left = 293, width = 250), page = 1, replace = FALSE)

Error Traceback:

17. FUN(X[[i]], ...)
18. lapply(X = X, FUN = FUN, ...)
19. sapply(boxes, function(box) {
    if (position["top"] > box["top"] && position["top"] < box["bottom"] &&
    position["left"] > box["left"] && position["left"] <
    box["right"]) { ...
20. FUN(X[[i]], ...)
21. lapply(xml2::xml_find_all(node_returned, xpath = "//text"), function(text_node) {
    position <- xml2::xml_attrs(text_node)
    position <- setNames(as.integer(position), names(position))
    boxed <- sapply(boxes, function(box) { ...
22. self$drop_unboxed_text_nodes(node = page_node, box_list, copy = TRUE)

.fun(piece, ...)
10. (function (i)
{
piece <- pieces[[i]]
if (.inform) { ...
9. loop_apply(n, do.ply)
8.
llply(.data = pieces, .fun = .fun, ..., .progress = .progress,
.inform = .inform, .parallel = .parallel, .paropts = .paropts)
7. dlply(.data = df, .variables = .(box_id), .fun = function(df2) {
page_node <- xml_new_root(page_nodes[[df2[["page_node"]]]],
.copy = TRUE)
box_list <- lapply(1:nrow(df2), function(j) setNames(as.numeric(df2[j, ...
6. .fun(piece, ...)
5. (function (i)
{
piece <- pieces[[i]]
if (.inform) { ...
4. loop_apply(n, do.ply)
3. llply(.data = pieces, .fun = .fun, ..., .progress = .progress,
.inform = .inform, .parallel = .parallel, .paropts = .paropts)
2. dlply(.data = self$boxes, .variables = .(page_node), .fun = function(df) {
df[["box_id"]] <- 1:nrow(df)
page_no <- unique(df[["page_node"]])
dlply(.data = df, .variables = .(box_id), .fun = function(df2) { ...

1. P$get_text_from_boxes(paragraphs = FALSE)

[ablaette]

Lieber Christoph – auf Anhieb kann ich es nicht einordnen, aber es sollte machbar sein … Ich komme hoffentlich morgen, sonst beim Rückflug dazu, die Sachen anzusehen. Dass man ein Verfahren für unterschiedliche Layouts auf der linken und der rechten Seite braucht, war eigentlich absehbar. Das bisschen blöde ist, dass man leicht den Orientierungssinn im Dokument verliert, wenn man anfangs Seiten am Anfang wegschneidet, d.h. bei den boxes definierte Seiten beziehen sich auf die XML-Knoten für Seiten, nicht auf die ursprüngliche Seitennummerierung… Wird schon gehen (-; Herzlicher Gruss, Andreas Von: ChristophLeonhardt <notifications@github.com> Antworten an: PolMine/trickypdf <reply@reply.github.com> Datum: Mittwoch, 9. Mai 2018 um 18:42 An: PolMine/trickypdf <trickypdf@noreply.github.com> Cc: Subscribed <subscribed@noreply.github.com> Betreff: [PolMine/trickypdf] P$get_text_from_boxes(paragraphs = FALSE) in seemingly normal document (#4) After parsing 142 optically identical documents, get_text_from_boxes throws following error: Fehler in if (position["top"] > box["top"] && position["top"] < box["bottom"] && : Fehlender Wert, wo TRUE/FALSE nötig ist Document can be found online here: https://www.landtag-saar.de/Plenarprotokoll/PlPr15_041.pdf Used box coordinates: P$add_box(box = c(top = 105, height = 690, left = 45, width = 250), page = NULL, replace = TRUE) P$add_box(box = c(top = 105, height = 690, left = 297, width = 250), page = NULL, replace = FALSE) P$add_box(box = c(top = 86, height = 700, left = 45, width = 250), page = 1, replace = TRUE) P$add_box(box = c(top = 86, height = 700, left = 293, width = 250), page = 1, replace = FALSE) Error Traceback: 1. FUN(X[[i]], ...) 2. lapply(X = X, FUN = FUN, ...) 3. sapply(boxes, function(box) { if (position["top"] > box["top"] && position["top"] < box["bottom"] && position["left"] > box["left"] && position["left"] < box["right"]) { ... 4. FUN(X[[i]], ...) 5. lapply(xml2::xml_find_all(node_returned, xpath = "//text"), function(text_node) { position <- xml2::xml_attrs(text_node) position <- setNames(as.integer(position), names(position)) boxed <- sapply(boxes, function(box) { ... 6. self$drop_unboxed_text_nodes(node = page_node, box_list, copy = TRUE) 7. .fun(piece, ...) 10. (function (i) { piece <- pieces[[i]] if (.inform) { ... 9. loop_apply(n, do.ply) 8. llply(.data = pieces, .fun = .fun, ..., .progress = .progress, .inform = .inform, .parallel = .parallel, .paropts = .paropts) 7. dlply(.data = df, .variables = .(box_id), .fun = function(df2) { page_node <- xml_new_root(page_nodes[[df2[["page_node"]]]], .copy = TRUE) box_list <- lapply(1:nrow(df2), function(j) setNames(as.numeric(df2[j, ... 6. .fun(piece, ...) 5. (function (i) { piece <- pieces[[i]] if (.inform) { ... 4. loop_apply(n, do.ply) 3. llply(.data = pieces, .fun = .fun, ..., .progress = .progress, .inform = .inform, .parallel = .parallel, .paropts = .paropts) 2. dlply(.data = self$boxes, .variables = .(page_node), .fun = function(df) { df[["box_id"]] <- 1:nrow(df) page_no <- unique(df[["page_node"]]) dlply(.data = df, .variables = .(box_id), .fun = function(df2) { ... 1. P$get_text_from_boxes(paragraphs = FALSE) — You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub<#4>, or mute the thread<https://github.com/notifications/unsubscribe-auth/AGg7NZ14a87zDrhT_VZrF58pSXVsW9obks5twxyGgaJpZM4T4pCZ>.