Added test for issue #2608.

Only run tests on recent mupdf which have fixed the underlying bug.

Test failed in optimised rebased; have fixed.

src/extra.i

@@ -3333,7 +3333,7 @@ void JM_append_rune(fz_buffer *buff, int ch)
     else
     {
         // 8 hex digits
-        snprintf(text, sizeof(text), "\\u%08x", ch);
+        snprintf(text, sizeof(text), "\\U%08x", ch);
         mupdf::ll_fz_append_string(buff, text);
     }
 }

tests/resources/test_2608_expected

@@ -0,0 +1,10 @@
+No significant gamma-ray excess above the expected background
+is detected from the direction of FRB 20171019A, with 52 gamma
+candidate events from the source region and 524 background event.
+A second analysis using an independent event calibration and reconstruction (Parsons & Hinton 2014) confirms this result. A search for
+variable emission on timescales ranging from milliseconds to several minutes with tools provided in (Brun et al. 2020) does not reveal
+any variability above 2.2 𝜎. For the total data set of 1.8 h, 95% confidence level (C. L.) upper limits on the photon flux are derived using
+the method described by Rolke et al. (2005). The energy threshold
+of the data is highly dependent on the zenith angle of the observations. For these observations, the zenith angles range from 15 to 25
+deg, which leads to an energy threshold for the stacked data set of
+𝐸th = 120 GeV. The upper limit on the Very High Energy (VHE)

tests/test_font.py

@@ -2,7 +2,7 @@
 Tests for the Font class.
 """
 import fitz
-
+import os
 
 def test_font1():
     text = "PyMuPDF"
@@ -37,3 +37,18 @@ def test_fontname():
         if str(e).startswith("bad fontname chars"):
             detected = True  # illegal fontname detected
     assert detected
+
+def test_2608():
+    if fitz.mupdf_version_tuple <= (1, 23, 4):
+        print( f'Not running test_2608 because mupdf too old: {fitz.mupdf_version_tuple=}')
+        return
+    flags = (fitz.TEXT_DEHYPHENATE | fitz.TEXT_MEDIABOX_CLIP)
+    with fitz.open(os.path.abspath(f'{__file__}/../resources/2201.00069.pdf')) as doc:
+        page = doc[4]
+        blocks = page.get_text_blocks(flags=flags)
+        text = blocks[10][4]
+        with open(os.path.abspath(f'{__file__}/../test_2608_out'), 'wb') as f:
+            f.write(text.encode('utf8'))
+        with open(os.path.abspath(f'{__file__}/../resources/test_2608_expected'), 'rb') as f:
+            expected = f.read().decode('utf8')
+        assert text == expected