Recently found out that the regex macro \d is not actually the same as [0-9], because, depending on the implementation, it might also decide to include every single number from all of unicode, which includes, but is not limited to:
0٠۰߀०০੦૦୦௦౦೦൦๐໐０
1١۱߁१১੧૧୧௧౧೧൧๑໑１
2٢۲߂२২੨૨୨௨౨೨൨๒໒２
3٣۳߃३৩੩૩୩௩౩೩൩๓໓３
4٤۴߄४৪੪૪୪௪౪೪൪๔໔４
5٥۵߅५৫੫૫୫௫౫೫൫๕໕５
6٦۶߆६৬੬૬୬௬౬೬൬๖໖６
7٧۷߇७৭੭૭୭௭౭೭൭๗໗７
8٨۸߈८৮੮૮୮௮౮೮൮๘໘８
9٩۹߉९৯੯૯୯௯౯೯൯๙໙９��
I have decided that from now on I am never using \d ever again.
List stolen from this stack overflow thread