Support inline formatted strings

DESCRIPTION

@@ -1,6 +1,6 @@
 Package: tidyxl
 Title: Read Untidy Excel Files
-Version: 1.0.9
+Version: 1.0.10
 Authors@R: c(
     person("Duncan", "Garmonsway", email = "[email protected]", role = c("aut", "cre")),
     person("Hadley", "Wickham", role = c("ctb"), comment = "Author of included readxl fragments"),

NEWS.md

@@ -1,3 +1,8 @@
+# tidyxl 1.0.10
+
+* Fixed a bug in the support for formatted strings, which sometimes weren't
+  being imported (#92)
+
 # tidyxl 1.0.9
 
 * CRAN fixes

src/xlsxcell.cpp

@@ -71,12 +71,12 @@ void xlsxcell::cacheValue(
     ) {
   // 'v' for 'value' is either literal (numeric) or an index into a string table
   rapidxml::xml_node<>* v = cell->first_node("v");
+  rapidxml::xml_node<>* is = cell->first_node("is");
+
   std::string vvalue;
   if (v != NULL) {
     vvalue = v->value();
     SET_STRING_ELT(book.content_, i, Rf_mkCharCE(vvalue.c_str(), CE_UTF8));
-  } else {
-    book.is_blank_[i] = true;
   }
 
   // 't' for 'type' defines the meaning of 'v' for value
@@ -100,15 +100,18 @@ void xlsxcell::cacheValue(
 
   if (t != NULL && tvalue == "inlineStr") {
     book.data_type_[i] = "character";
-    rapidxml::xml_node<>* is = cell->first_node("is");
     if (is != NULL) { // Get the inline string if it's really there
+      // Parse it as though it's a simple string
       std::string inlineString;
       parseString(is, inlineString); // value is modified in place
+      // Also parse it as though it's a formatted string
       SET_STRING_ELT(book.character_, i, Rf_mkCharCE(inlineString.c_str(), CE_UTF8));
+      book.character_formatted_[i] = parseFormattedString(is, book.styles_);
     }
     return;
   } else if (v == NULL) {
     // Can't now be an inline string (tested above)
+    book.is_blank_[i] = true;
     book.data_type_[i] = "blank";
     return;
   } else if (t == NULL || tvalue == "n") {

tests/testthat/test-inline-strings.R

@@ -9,6 +9,16 @@ test_that("can read sheets with inlineStr", {
   expect_equal(x[14], "RQ11610")
 })
 
+test_that("can read sheets with formatted inline strings", {
+  # Original source: http://our.componentone.com/wp-content/uploads/2011/12/TestExcel.xlsx
+  # These appear to come from LibreOffice 4.2.7.2.
+  cells <- xlsx_cells("inline-formatted-string.xlsx")
+  index <- 64
+  expect_equal(cells$is_blank[index], FALSE)
+  expect_equal(cells$data_type[index], "character")
+  expect_failure(expect_equal(cells$character_formatted[index], NULL))
+})
+
 test_that("does not crash on phonetic strings", {
   # https://github.com/nacnudus/tidyxl/issues/30
   expect_error(xlsx_cells("phonetic.xlsx"), NA)