I'm not normally a fan of pastry, but I'll make an exception for certain Chinese pastries, particularly these light, deep-fried puffs filled with shredded daikon. The Chinese name is 蘿蔔絲酥餅 (luó bo sī sū bǐng), though you might also see this abbreviated as 蘿蔔酥 (luó bo sū). 蘿蔔（luó bo) is daikon/mooli/Chinese radish (though as discussed before, the term also covers a few other root vegetables), 絲 (sī) means "shredded" (referring to the 蘿蔔), 酥 (sū) means "crispy" (referring to the pastry), and 餅 (bǐng) denotes a biscuit/cake sorta thing.
English translations I've seen for 蘿蔔絲酥餅 include "deep fried turnip puff pastry", "crispy turnip puff pastry", "shredded turnip puff pastry", "crispy shredded turnip", and, slightly bizarrely, "mooli croissant". They're sometimes available in vegetarian versions (for example at Shanghai Blues in London), but if they're not explicitly marked as vegetarian, there may be lard in the pastry and/or pork mince in the filling.
Like European puff pastry, the pastry used to make 蘿蔔絲酥餅 consists of multiple layers which separate and flake up on cooking. However, instead of the layers being separated by pats of butter, they're separated by a rich, lard-heavy dough — you essentially make two doughs, one including water and one not, and layer them up, then fold and reroll a number of times to increase the number of layers. Another difference is that the folding process aims to expose the edges of the dough layers, so when the pastries are cooked they make a pretty pattern as shown in the photograph above (which was taken at Gerrard's Corner in London Chinatown).
I looked at a few different recipes when making these: Sunflower's recipe, Lily Ng's recipe, and two Red Cook recipes for beet puffs and durian puffs. All of these make different quantities and use different amounts of flour, water, and fat — and Sunflower's recipe substitutes oil for part of the lard, while Lily's recipe adds an egg to the water pastry. I thought the best way to figure out what to do was to work out the ratios of the ingredients by weight, and compare these ratios between the recipes.
I eventually settled on a fat:flour:water ratio of 30g:100g:40g for the water pastry and a fat:flour ratio of 60g:100g for the lard cake — this fitted pretty closely to Sunflower's and Lily's recipes (which were given in weights), and also to one of the Red Cook recipes (the beet puff ratios were very different to the other three, and I do wonder if the use of volume measurements may have led to inaccuracies). I made one batch using all lard, and one batch using Sunflower's suggestion of replacing some of the lard with oil. The eventual flavour wasn't noticeably affected by the lesser quantity of lard.
Regarding methods, there seem to be two main options: either treat the pastry as a whole, and repeatedly fold and roll the two doughs together before cutting into pieces against the layers (Lily's recipe), or divide each dough into portions and combine them individually (the other three recipes). I tried both, using Lily's method with the all-lard dough and the other method (as described by Sunflower) with the lard-oil dough. I found that Lily's method was much less faff, but the other method produced better results for me, with the flaky strips more apparent. I don't know how much of this was down to the difference in dough composition and how much to the difference in method. (Edit, April 2011: here's an illustrated guide to different folding methods.)
Finally, there are also two options for cooking the things — bake them at 200°C (400°F) for around 20 minutes, or deep-fry them. I cooked half of each batch with each method. Unsurprisingly, the deep-fried ones were flakier while the baked ones were more solid. The all-lard baked ones ended up lighter in colour than the lard-plus-oil baked ones, but that might have been partly because they went in colder, due to the chilling of the dough. The all-lard fried ones were darker, denser, and less flaky than the lard-plus-oil fried ones, but that might have been partly or entirely because I fried them second and the oil was hotter and already had bits in (which can speed up browning).
The main mistake I made was in not putting enough filling in — I was worried that they'd come apart, but in the end only one or two of them leaked slightly. Next time I'll roll the pastry thinner and add more filling.